Sub-70 half marathon training using a percentage-based approach

One big advance for me as a coach in the last few years has been the “feel” that I have for half marathon training. I’m now at a point where I have a good sense of how to design a training program to produce reliable results in the half marathon for a wide range of different kinds of athletes: men, women, elite, subelite, amateur, etc. 

There is, of course, not one singular HM training plan that’s perfect for everyone, but there are principles of half marathon training that work across different athletes.

The half marathon is an odd distance because very few books or training programs explicitly focus on it as a “first class” racing distance—usually, the advice you get is to do a generic 10k plan but with more endurance, or a generic marathon plan but with more speed.

My goal here is to show how Canova-style “full-spectrum” percentage-based training can be applied to building training plans targeting the half marathon, using worked examples from four different athletes I have coached who were all aiming for the same goal: a sub-70 (i.e. sub-1:10) half marathon. 

📅 If you just want to see the HM training calendars, click here!

If you are interested in marathon training, I have a brand-new book on exactly that subject! Check out my new book, Marathon Excellence for Everyone, if you want to see how I use the same Canova-style "full-spectrum" approach to train marathoners.

Looking at real schedules for lessons about real training

Even if you are aiming for much slower than 1:10 (or much faster!) there’s a lot to learn from looking at different training plans organized around the same goal, so don’t worry too much about the exact performance level: we’ll be using percentage-based workouts anyways, so it will be very easy to convert workouts to different performance levels.

This article will take a slightly different approach than my full-spectrum percentage-based 10k training plan article—instead of building up a generic training schedule for a hypothetical athlete, we’ll look at four examples of real training schedules that led to success in the half marathon.

Looking at real training will give some insights into the kind of individualization that’s necessary for a good half marathon program, given an athlete’s background, strengths, weaknesses, and life situation. Often I feel like articles on training (including my own!) talk about the importance of individualization but don’t go in-detail about how you would actually individualize different aspects of the plan, and why.

Presenting multiple schedules simultaneously also helps to establish what principles work across individuals. It’s risky to base your training on something that worked for just one specific person—you don’t know how much of their success is due to the specifics of the schedule, the principles behind it, or simply their talent level.

Meeting our sub-70 half marathon athletes

We’ll be looking at training from four different athletes that I currently coach. In all cases, they were training for a half marathon as their main goal race for the season, not doing it en route to racing a marathon.^[1] All of these training programs occurred in the last two calendar years (2024–2025) and culminated in a half marathon personal best by at least 20 seconds.

For anonymity I’m rounding all of their times to the nearest 10 seconds and in some cases I am not mentioning specific aspects about their life that may have affected the individualization of the training schedule (e.g. personal life or job issues). All of these athletes are men, since I have not (yet!) coached a woman to a sub-1:10 half marathon. 

None of these runners did particularly high mileage for their HM buildups—all ran between 70 and 90 mi/wk (though some of them have done more mileage in the past, usually for marathon training), and all of them work full-time jobs.^[2]

Keeping that in mind, here’s a quick intro to our four case studies.

Athlete	Age	Short bio	HM result
Athlete A	Late 20s	Did not compete in high school or college. Started running five years ago, initially focused on a mix of trail, road, and ultra events.	1:08:20
Athlete B	Late 30s	Competed in high school and college (Division 3). >10 years of road running experience, mostly marathon-focused.	1:09:00
Athlete C	Mid 30s	Did not compete in high school or college. Started running recreationally in late 20s, ~5 years of competitive running experience.	1:08:40
Athlete D	Mid 20s	Competed in high school and college (Division 1). ~10 years of competitive running experience, mostly focused on 1500–5000 m.	1:07:20

Already, it’s easy to see why considering an athlete’s “training age” is important – Athlete D is about ten years younger than Athlete C, but has many more years of experience with serious training!

From these short bios, we can also get a sense of how an athlete’s background will inform their training approach.

For instance, Athlete A’s experience doing trial and ultra races means he has a very good base of “long endurance” and needs to spend less time doing long easy runs and other general endurance work to build out support for half marathon specific workouts.

Athlete D, on the other hand, is younger and until fairly recently was doing 1500/5k focused training, so he will be much more comfortable with fast workouts.

High-level principles for sub-70 half marathon training

Before we get into the schedules—which may look pretty sophisticated at times—I want to emphasize that the principles of half marathon training are very simple.

Training is divided into three phases:

1. A general phase focused on increasing volume and improving capabilities across a wide range of speeds, with a particular focus on high-end aerobic workouts.
2. A race-supportive phase focused on building speed at 105–110% HMP and endurance at 90–95% HMP
3. A race-specific phase focused on long, fast workouts at half marathon pace (100% HMP), supported by high volume sessions at 95% and 105% HMP, all with plenty of recovery afterwards.

As for key HM-specific workouts, there are really only three types of sessions the athlete needs to be able to run in order to run a fast half marathon.^[3] These are:

1. Long fast runs at ~95% HMP, ideally building to 13–15 mi / 20–25 km continuous
2. Long repeats at 100% HMP with 1 km “floating” recovery at 85–90% HMP, ideally building up to ~15 km total at HMP, individual repeats of 2–4 km, and a total workout volume of ~20–22 km
3. Long repeats at 105% MP, ideally building to ~10 km total volume with individual repeats of 2–3 km in length

With real schedules, the questions are (a), how to work towards these goals under the constraints of athlete’s life and current fitness, and (b) what work needs to be done in the race-supportive and general phases to enable these key training sessions (since you can’t just jump into 15mi at 95% HMP right way).

So, when translating these principles to a real athlete, the old quote applies: “simple ain’t easy.”

Phases of training for the half marathon

As a reminder, in a full-spectrum percentage-based system, training is divided into three phases: the general phase (sometimes the “Fundamental” phase), the race-supportive phase (sometimes the “Special” phase), and the race-specific phase.

In an ideal world, the race-supportive and race-specific phases would each be 6–8 weeks long, and the general phase would be 8–12 weeks long. In reality, we (meaning me plus the athlete I’m working with) have to work around the constraints of the calendar—competitive opportunities for sub-70 HM runners only come around a few times a year, so we’re constrained by when our goal race is happening, and often also by when the last goal race happened.

Here’s what the considerations around periodization looked like for each runner:^[4]

Athlete	Prep time	Notes
Athlete A	11 weeks	HM race was 12 weeks after a marathon; we took one week “off from training” post-marathon for recovery (but still included some easy running).
Athlete B	20 weeks	Was starting up with training again after a period of a few months of unstructured running.
Athlete C	8 weeks	HM race was 8 weeks after a marathon; like Athlete A we took one week “off from training” post-marathon.
Athlete D	11 weeks	Was already training consistently at ~50 mi/wk with one threshold workout per week.

 Now, what does individualization look like here? Athletes A and C need a few weeks of “regenerative” or “introductory” training to build back elements of general fitness that we (intentionally) neglected during the final weeks of their marathon buildup. This regenerative training also needs to help restore their mental energy and enthusiasm for training—so we should not use long or strictly-paced sessions for the first several weeks of training.

Athlete B has been away from structured workouts for a while, so he will need a longer traditional “base” with a long general phase. He can also afford to use a long general phase because he has many weeks to prepare.

Athlete D is in the middle: he is already doing some general training but because of his middle distance background, still needs more general training than Athletes A and C.

Introductory training versus general training

Let’s talk briefly about returning to the general phase after concluding your previous season. When you are fresh off a big race and are starting up a new training cycle, you need a “bridge” back to traditional general-phase training. Relying on Canova’s terminology, this is “introductory training.”

As I alluded to above, you don’t want to just start up with generic base training without considering your recent race-specific training—if you are transitioning into HM training coming off a 5k-focused season, your first few weeks of the general phase should look different than if (like Athlete A and Athlete C) you are starting up immediately after a marathon.

So, it helps to further subdivide the general phase into a few weeks of introductory training, if needed, then the remainder of the general phase.

The goals are similar, but the mentality is different: the introductory phase focuses on re-introducing aspects of training that were missing during the final weeks of the last training program—things like speed and short hill repeats for a marathoner, and long easy runs and steady running over cross-country terrain for a 1500m/5k runner.

Restoring mental energy is of paramount importance during the introductory period. Workouts are primarily effort-based, not strictly paced, and we go out of our way to do training sessions that are enjoyable and invigorating, as opposed to draining.^[5]

Athletes B and D are already at a steady baseline of training and do not need much of an introductory period; they can go right into general training.

Visualizing half marathon periodization

The first things to go on the training calendar are the half marathon itself, plus any lead-up competitions. Then, we divide out the calendar into our phases of training.

Here’s a visual representation of how I sketched out the phases of training for each athlete, with all competitions indicated on the schedule as well:

While these plots look nice and "clean," remember that phases of training blend smoothly into one another; it’s not like we drop everything and completely change workouts when shifting to a new phase of training.

Planning out key workouts for half marathon performance

At a conceptual level, here’s how I think about planning out key workouts:

1. The half marathon will require 13.1 mi / 21.1 km of fast running
2. What are the key HM-specific workouts that enable this performance?
3. What are the key HM-supportive workouts that enable the HM-specific workouts?
4. What are the aspects of general fitness that enable the HM-supportive workouts?

I say “at a conceptual level” because at a practical level, we often don’t get to fully “work in reverse” like this—we have to start from where the athlete actually is, and we may have to get creative in terms of how we build up the support we need to run 13.1 miles at half marathon pace.

Half marathon specific workouts: 95–105% HMP

The most important training sessions for half marathon performance are workouts within five percent of half marathon pace (HMP). We can categorize these sessions as:

HM Specific Endurance: Long fast runs and/or stepwise long fast runs at 95% HMP (this is roughly marathon pace) of anywhere from 10–16 mi / 16–26 km

HM Specific: Long repeats of 1–4 km, often with “floating” recovery, at 100% HMP

HM Specific Speed: Medium to long repeats of 1–4 km, and sometimes short fast continuous runs or competitions, at 105% HMP (this is roughly 8k–10k pace)

Remember you can easily calculate these workout paces using my handy pace percentage calculator!

Guidelines for HM-specific workouts

For a “generic” sub-1:10 HM runner, here are some approximate guidelines for what I’d ideally like to see for their final, biggest sessions at each of these paces: 

Workout	Guideline
95% HMP	13–15 mi / 20–25 km at 95% HMP, or a step-wise progressive session such as 5-5-5 mi (or 8-8-8 km) at 92-94-96% HMP
100% HMP	14–15 km of repeats of 2–3 km at 100% HMP with a 1 km float recovery at 90% HMP, for a total session volume of ≈20 km
105% HMP	10 km of repeats of 2–3 km at 105% HMP, plus either a competition of 8–10 km or a fast continuous run of 6–8 km at 105% HMP in training

Again, “ideally” is doing a lot of work here! You will see below that we do not actually hit all of these goals with the actual real athletes.

Half marathon supportive workouts: 90% HMP and 110% HMP (nominally)

To support the HM-specific work, we want strong endurance at 90% HMP and strong speed at 110% HMP.

HM-supportive speed: Repeats of 400–1000m or fartlek repeats of 1–3 min at 107–110% HMP off a moderate “floating” recovery.

This pace is slightly on the optimistic side of 5k pace according to a typical conversion chart, but you need to be careful because 110% HMP is fast enough that endurance-oriented vs. speed-oriented HM runners will have very different limitations 110% HMP

HM-supportive endurance: Long fast runs at 90% HMP, or stepwise progressive sessions at 85% progressing to 92% HMP.

Individualization is especially important with HM-supportive speed; of the four athletes above, their actual 5k PR pace is anywhere from 105% HMP to 110% HMP (calculated using their HM result after this training cycle). On race day when they are in peak HM shape, their current 5k fitness is probably more like only 104–108% HMP!

Often I’ll just consider “110% HMP” as a nominal label, and in actuality will use my best estimate of their current 5k fitness as the basis for their HM-supportive speed workouts. This same individualization applies to HM-specific speed at 105% HMP though to a lesser extent; you can similarly just call those workouts “8k/10k pace” and work off a fitness estimate for those paces.

In practice you can often combine workouts for HM-supportive endurance (90% HMP) with workouts for HM-specific endurance (95% HMP), for example using a stepwise progression run from 90–95% HMP.

This type of combination is a good solution for shorter buildups where you don’t have enough time to build up long runs at 90% HMP, then build up long runs at 95% HMP afterwards.

Sometimes it’s also possible to leap over long runs at 90% HMP by taking shorter runs at 95% HMP and just extending them up until you get to the requisite volume of HM-specific endurance—a solution that works well for runners strong at shorter distances who are moving up to the half marathon. You still want to touch on 90% HMP sometimes, though: one good way to do this is with shorter steady runs of 4-6 mi / 6-10 km as a second run for the day on an otherwise-easy day.

Guidelines for HM-supportive workouts

Again, here are guidelines for a “generic” sub-1:10 HM runner in terms of what I’d like to see for the final biggest sessions at these paces:

Workout	Guideline
110% HMP	35–45 min of mixed fartlek off a moderate pace, e.g. 3 sets of 4-3-2-1 min at 105 → 110% HMP with 1 min moderate float at 85–90% HMP; also, in a different session, 6–7 km of repeats of 400–1000 m at 107–110% HMP
90% HMP	15–18 mi / 25–30 km at 90% HMP, or a step-wise progressive run like 5-5-5 mi at 86-88-92% HMP

The mixed fartlek sessions are an effective way to improve lactate shuttling and oxidation, and they also serve well as speed support for alternating kilometer workouts at 100% HMP.

General phase workouts for the half marathon

Our goal in the general phase is to build up a broad base of fitness across a wide range of speeds, while steadily increasing overall training volume (both in workouts and in terms of weekly mileage).

For the half marathon I like to ask “what sessions will support the HM-supportive and HM-specific workouts necessary for a good race performance?”

In light of that question, here’s a table showing what kinds of general phase workouts support each of the types of HM-specific and HM-supportive sessions we saw above:

HM workout	General-phase support
Long fast run at 90% HMP	Long easy runs through rolling hills, up to two hours
Long fast runs at 95% HMP	Long moderate runs at ~80–85% HMP up to 90 minutes and sometimes through rolling hills Strong runs at ~95% HMP of 6–9 mi / 10–15 km or 30–50 min Long runs of 70–90 min with a progressive finish over the last 20–45 min
HM-specific repeats at 100% HMP	Classical threshold (LT2 / SSmax) work Double threshold days Long sub-threshold repeats Shorter continuous runs at ~98–102% HMP Kenyan-style progression runs
HM-specific speed at 105% HMP	Repeats of 600–1000 m at 105% HMP Longer repeats of 4–8 min at 8k / 10k effort over cross-country terrain or through rolling hills Medium- to long repeats of 3–8 min up a steady incline Fast continuous runs of 12–20 min or 4–7 km at 105% HMP (sometimes over cross-country terrain, rolling hills, or on a steady incline)
HM-supportive speed at 110% HMP	Short uphill repeats of 1–2 min Short fartlek repeats of 1–2 min Cut-down repeats of 1–3 min starting at HM effort and working down to faster-than-5k pace Short fast continuous runs of ~3–4 km or 10–15 min

Another perspective on the general phase might be “what general aspects of fitness are required for the half marathon?” That question leads to a few additional components of fitness to make sure we target during general training:  

Aspect of fitness required for HM	How to build during general phase
Ability to cover 13 mi / 21 km without body breaking down	High overall mileage Some runs of 13+ mi / 21+ km
High VO2max	Interval workouts ~105–110% HMP
High SSmax	Interval work at 105% HMP Classical threshold (LT2) workouts Sub-threshold workouts Double threshold days
Good running economy	High overall mileage High-volume workouts at 95–105% HMP Hill sprints
Good mechanics at high speeds	Hill sprints Strides

Clearly, we have many options for building all-around fitness in the general phase, and we do not need to use all of these workouts. A wide range of different approaches can work during this phase of training, as long as they use high overall volume and a healthy amount of high-end aerobic running.

Planning out key half marathon workouts on the training calendar

When I’m actually coaching, I don’t write out the exact workouts any further out than two or three weeks. For example, I know that we need to do some kind of HM-specific session at 100% HMP about 10–15 days before the race, and this should be our biggest / most focused session of that type, but the specific parameters will depend a bit on how the training buildup goes.

When I write out hypothetical schedules it is much easier. For example, here is a very standard progression of HM-specific workouts:

Weeks before HM	Workout	Parameters
6	10 × (1 km at 100% HMP, 1 km at 85–90% HMP)	20 km total 10 km at HMP 1 km extension avg. pace 95% HMP
4	7 × (2 km at 100% HMP, 1 km at 85–90% HMP)	21 km total 14 km at HMP 2 km extension avg. pace 96.7% HMP
2	5 × (3 km at 100% HMP, 1 km at 85–90% HMP)	20 km total 15 km at HMP 3 km extension avg. pace 97.5% HMP

An athlete who is healthy (no injury, no sickness) and successfully completes this progression of workouts—not in isolation, but as a capstone to the other components of training—is basically guaranteed to run a strong half marathon. Indeed, it’s easy to see how the final destination of this progression is just 21 km at 100% HMP.

However, the actual reality of coaching can end up being very different, and we have to dynamically adapt the later workouts based on how the early season workouts go (not just at this pace, but across all paces). The goal needs to be running fast for 13.1 mi / 21.1 km, not adhering to a predetermined training plan.^[6]

The same goes for all the other types of workouts. The high-level goal is to emphasize workouts at 95–105% HMP during the race-specific phase, workouts at 90–110% HMP during the race-supportive phase, and a wide range of different fundamental workouts during the general phase. But where the workouts actually end up on the calendar depends a lot on how each athlete’s real life and real training play out.

Mileage: around 70–85 mi/wk / 110–135 km/wk

The half marathon is the longest race distance where it’s practical to have your mileage “make the race distance manageable” in the sense of having 13 miles not be an inordinately large stress on your body. Under that view, without injury or scheduling constraints, 90 mi/wk is a good bar to shoot for because 13 miles per day is 91 miles per week.

Still—and this might be a bit iconoclastic—I have found that having very high mileage is not absolutely essential for the half marathon, at least at the sub-elite level (maybe it is a different story for a sub-60 half marathon…).

Running sub-1:10 on ~70 mi/wk (~110 km/wk) is very doable for runners of moderate talent if their training is good. I would like to see mileage of more like 80–90 mi/wk (130–145 km/wk) if possible, and would love 95+ mi / 150+ km, but we can pretty comfortably make things work on 70 mi/wk / 110 km/wk or even a bit less in some cases.

With that in mind, here is the actual weekly mileage for our four sub-70 HM runners during their buildups:

Keep in mind that in the final 2–4 weeks, mileage is less important than getting in high-quality race-specific workouts, and recovering well afterwards. After big HM-specific sessions, we keep things as short and as easy as we need to in order to recover well.

Basic mileage paces: very easy, easy, and moderate run paces

Under a full-spectrum view of training, there is no such thing as a singular “easy run pace.” Some runs are very easy regeneration runs, some runs are easy, and some are moderate to strong. These four sub-1:10 HM runners typically run 6:45–7:30/mi / 4:15-4:40/km on a “normal” non-workout day, but it can vary a decent amount, and as long as they are able to run well in workouts and recover afterwards, I’m not super concerned about how fast or slow they run on any particular day.

That said, here are some general guidelines for very easy, easy, moderate, and strong paces for effort-based runs, given your current HM fitness:

Effort	Guideline
Very easy	55% HMP or slower
Easy	60–70% HMP
Moderate	75–85% HMP
Strong	85–95% HMP

Full training calendars for sub-70 half marathon runners

We’re now at a point where we can look at the full training logs. I’ve posted links to PDF downloads below.

Note that these are not the training plans as-received by the athlete; they get a much more detailed schedule (like this one) with in-depth workout instructions, pace ranges, effort guidelines, etc.

📅 Athlete A: 1:08:20 HM on 70–80 mi/wk / 110–130 km/wk || absolute pace version (in /mi & /km)

📅 Athlete B: 1:09:00 HM on 80–90 mi/wk / 130–145 km/wk || absolute pace version (in /mi & /km)

📅 Athlete C: 1:08:40 HM on 65–70 mi/wk / 105–110 km/wk || absolute pace version (in /mi & /km)

📅 Athlete D: 1:07:20 HM on 60–70 mi/wk / 95–110 km/wk || absolute pace version (in /mi & /km)

As usual I’ve provided these as relative paces (% HMP) and also absolute paces (per mile and per km); however please keep in mind that relative paces, especially in long training cycles, are more of an art than a science.

Athlete B is a good example; my underlying estimate of his current fitness changed a lot over the course of 20 weeks, so seeing his progress is actually much easier in the absolute pace version.^[7]

Note there are a few things that we did in training that are not on these calendars, namely strides, mid-run pickups, post-run strength routines, and in some cases hill sprints.

In the final part of this article we’ll look at the progression of each type of workout, and take a closer look at the ways I individualized training for each athlete.

Individualizing introductory and general phase training for the half marathon

The general phase, and any introductory phase that accompanies it, is highly amenable to individualization; as mentioned earlier, many strategies can work during this phase of training. Here’s how I approached the general phase for each athlete.

Athlete A: General phase

Athlete A was coming off a marathon buildup (where he ran a fast time on a flat course with good conditions), so we needed to focus on regeneration and re-integration of basic training elements for a solid month before we could move to anything HM-specific.

Since he just ran 26.2 miles at marathon pace, long fast runs at precise paces were not necessary during the first several weeks of training. Instead, we needed to rebuild aspects of fitness we had (intentionally) neglected during the final stages of marathon training, and we needed to do so in a way that was not mentally draining.

Effort-based progressive workouts on hills and over cross-country terrain are ideally suited for these goals. Here are the key introductory and general phase workouts I used for this athlete:

Key general phase workouts: Athlete A
7 mi Kenyan-style progression on XC terrain
3 sets of 3’-2’-1’-45”-30” fartlek at ~8k → ~3k effort with 1’ easy-run recovery
20’ fast progressive run on XC terrain, then 8 × 45” very fast uphill (8% grade)
4 × 6 min at 10k / 8k effort on XC terrain with 3’ jog
30’ easy + 30’ moderate + 30’ progressive on hilly dirt roads
10 × 1 km cut-down on XC terrain with 2’ jog

Athlete B: General phase

When Athlete B started up training again he was already running about 65 mi / 105 km per week of strictly easy running, and he had not competed in several months. So, he had little need for a transitional or restorative introductory phase—we could just start a gradual ramp-up of general fitness workouts.

The early part of his general phase looks pretty standard: many classical lactate threshold workouts, longer continuous tempo work, and a mix of shorter repeats, hills, and longer repeats aimed at improving his “aerobic power,” roughly connected with 5k/10k fitness.

One thing we did not use is very big long fast runs or very long easy runs. This athlete takes very well to long fast runs, so I knew we would be able to incorporate those later in training without a problem. The real challenge, as illustrated by his 27:10 8k race early on in training, was a lack of aerobic power.

We used a few double threshold days, and I had planned on using more, but between sickness and some travel, we were not able to use very many of these sessions, so we instead had to focus on a longer, slower build-up of work at 105–110% HMP to work the faster-than-HMP range.

Many of the early general phase workouts were purely effort based, in part because it had been a long time since this athlete had raced and also frankly because he was not in very good shape yet.

Here are some key sessions we used in the general phase to solve this problem:

Key general phase workouts: Athlete B
30 min moderate at ~88% HMP
12 × 2 min at ~8k effort with some progression, 45” jog
10’, 8’, 8’ at threshold effort with 2’ jog
14 mi with last 4 mi progressive
8k competition
10 × 2 min fast uphill with easy jog down
5 × 2’ fast uphill at 5k effort, 2 × 5’ on flat ground at 8k effort, 5 × 2’ at 5k effort
AM: 6 × 1k at threshold effort with 1’ walk / PM: 10 × 500 m at 10k effort with 30” walk
3 × 2 mi at threshold effort (~102% HMP) with 1.5 min jog
2 × (10 × 500 m at 105% HMP, about 10k effort, with 30” walk), 3–4’ walk/jog
AM: 6 mi at 95% HMP / PM: 4 × 5’ at 102% HMP with 1’ jog

Athlete C: General phase

Athlete C was, like Athlete A, fresh off a marathon, but had even less time to prepare. The focus for his first few weeks was regeneration after the marathon—we very much wanted to avoid strictly paced workouts for as long as possible, using effort-based sessions to rebuild fitness and also rebuild mental energy.

With how little time we had to prepare, we did practically no “real” general phase—we went immediately from introductory training into race-supportive and then race-specific training!

As with Athlete A, we had no need for long fast runs in the first several weeks of training—he had just run 26.2 miles at marathon pace, after all.

Here is a good test of your understanding of what we’ve discussed so far: can you identify why the following workouts were important for this athlete?

Key introductory workouts: Athlete C
70’ very easy on forest trails with friends
10 × 3 min at ~6:10/mi with 1 min walk (with fiancée)
6 mi strong at 93% HMP (with friends)
7 mi Kenyan-style progression run
10 × 2 min cut-down from controlled to optimistic 5k effort with 1 min walk

Athlete D: General phase

Athlete D was already running about 45–50 mi/wk (72–80 km/wk) and doing one workout each week (usually 4–5 x 1 mi at threshold effort with 1–2 min rest) and had not raced in a while, so like with Athlete B, we did not need a long introductory period.

Unlike Athlete B, he was already very strong in terms of speed—he was very comfortable running 4:40–4:50/mi / 2:55–3:00/km for short to medium-length repeats even at the start of training.

However, he also had some lingering hamstring soreness that would flare up when he ran fast or ran on hills, so we avoided shorter, faster workouts to a greater extent than I usually would until his hamstring issues abated.

With his background as a 1500/5k runner, and with this being only his second half marathon ever, the concern was always going to be the strength to run fast for a long time. In college, he did not usually do long fast runs, nor had he used float-recovery workouts.

So, his key general phase workouts were geared much more towards supporting his ability to, eventually, be able to run 13–15 miles at 95% HMP and successfully complete HMP-specific alternating kilometers, while retaining as much speed and aerobic power as possible without aggravating his hamstring.

Key general phase workouts: Athlete D
Increasing mileage to 60+ mi/wk / 100+ km/wk
16 mi easy long run
10 mi steady on flat ground at 85% HMP
8 × 1200 m at threshold effort with 1’ jog
3-2-1-3-2-1 k progressive sub-threshold (faster on shorter reps) with 2’ walk rest, ~96-98% HMP

Individualizing long fast runs at 90–95% HMP

Long fast runs are one of the cornerstones of a good half marathon. The real question is not whether to do them, but how to build up to the longest, fastest workout (which I usually schedule for about three weeks out from race day). Here's where we put long fast runs on the calendar for each athlete:

And here’s how I individualized the progression of long fast runs for each athlete:

Athlete A: Long fast runs
Arguably, marathon competition at 95% HMP
90’ as easy / moderate / strong
8 mi stepwise progression (2-2-2-2 mi) at 90-92-94-96% HMP
15 mi through hills at 87% HMP
8 mi at 94% HMP
10 mi through hills at 88% HMP (very hot/humid)
16 mi through hills at 83% HMP (very hot/humid)

As noted earlier, the “last” long fast run before this buildup for Athlete A was a marathon race! So, it would be totally silly to worry about doing long easy runs or long moderate runs early on in training. Athlete A “started from speed,” using short fast runs to eventually build up to long fast runs, using hilly routes to build strength.

He also trained in significant heat and humidity for the last two sessions, but I was not at all worried about the slow paces—his workouts at 100% HMP were excellent, and his shorter workouts at 94–96% HMP went well too. With that and his recent marathon performance, I had no concerns about his endurance.

Athlete B: Long fast runs
7 mi at 88% current HMP (84% eventual HMP) – actual pace 6:05/mi / 3:47/km
8 mi at 93% current HMP (90% eventual HMP) – actual pace 5:48/mi / 3:36/km
10 mi at 93% current HMP (90% eventual HMP) – actual pace 5:47/mi / 3:36/km
12 mi at 95% current HMP (92% eventual HMP) – actual pace 5:41/mi / 3:32/km
15 mi at 91% current HMP (88% eventual HMP) – actual pace 5:53/mi / 3:39/km
15 mi at 94% HMP – actual pace 5:38/mi / 3:30/km

Athlete B is a very experienced marathoner who takes extremely well to long fast runs. I knew that we could increase the distance of his long fast runs over time without any issues, as long as he had the aerobic power to support it.

In these examples I annotated the runs both with my estimate of his current HMP at the time of the workout, as well as the same pace expressed relative to his eventual HMP on race day, and the actual pace of the workout. This progression is an excellent example of what progress looks like: a combination of running faster at the same pace, running the same distance faster, and running fractionally slower for significantly longer.

Once we got up to 10 mi at ~93% current HMP, we “bifurcated” long fast runs into continuing to increase distance at the same effort level, and also adding a slower long fast run of 15 mi at ~90% HMP to help support the eventual move to 15 mi at 95% (which we nearly managed; in actuality we did 15mi at 94%).

Athlete C: Long fast runs
Arguably, marathon competition at 95% HMP
6 mi at 92% HMP
9 mi at 91% HMP (windy)
5-4-3-2-1 mi (15 mi / 24 km) stepwise progression at 90-93-96-99% HMP

Like Athlete A, this runner was already prepared for long, fast running at precise paces. With such a short buildup, we only did one true long fast run, letting his marathon fitness carry his endurance almost all the way to the end of the half marathon build.

Even that final stepwise progression was probably a little too early—it was tougher than I would have liked, and if we had a few more weeks to train I would have made this only 12 or 13 miles (~20 km).

Athlete D: Long fast runs
10 mi at 85% HMP
8 mi at 93% HMP
12 mi at 88% HMP
16 km (10 mi) stepwise 4-4-4-4 km at 91-93-95-97% HMP
13 mi at 95% HMP

Given Athlete D’s middle distance background, running fast was always going to be easy. To run long and fast, we needed to start with a foundation of shorter fast runs (8 mi at 93% HMP), running slower for a similar distance (12 mi at 88% HMP), doing part of a long fast run at 95%+ HMP (in the stepwise 16 km run), and finally the ultimate goal of 13 mi at 95% HMP.

Individualizing race-supportive speed (110% HMP, nominally)

I have fewer specific comments on race-supportive speed, since some of these athletes did not have enough time to do more than a couple sessions at this pace (or just one, for Athlete C!). Keep in mind the notes earlier on how workouts at “110% HMP” often end up being more like 107% HMP for endurance-oriented runners.

Recently, I have been using time-based fartlek workouts with a fast, floating recovery as supportive speed for the half marathon, with good success: I think the float recovery helps improve lactate shuttling and lactate oxidation capabilities, which translates very well to the half marathon (and also the marathon, since all four of these athletes are current or soon-to-be marathon runners).

One thing I did not use very much with these athletes is traditional track work, like fast 400m or 600m repeats.

Canova’s athletes, like Emile Cairess and Moses Mosop, will often do this kind of work even in marathon buildups, but I don’t have a strong sense of the right way to use these faster track sessions for longer races, especially when doing a short buildup.^[8]

So, while I’m planning on experimenting more with these types of track sessions in future buildups, you will not see very many of them in these half marathon schedules.

Note that some of these workouts were also mentioned in the key general phase workout section; that’s because general and supportive speed somewhat blur into one another, especially in shorter buildups.

Key HM-supportive speed workouts
Athlete A
3 sets of 3’-2’-1’-45”-30” fartlek at ~8k → ~3k effort w/ 1’ easy-run recovery
20’ fast progressive run on XC terrain, then 8 × 45” very fast uphill (8% grade)
8 × 2’ at 5k effort, 1’ easy + 8 × 1’ at 3k effort, 1’ easy, all on XC terrain
Athlete B
8 × 800 m at 104 → 108% HMP w/ 2–3’ jog
8 × 1 km cut-down at 103 → 106% HMP w/ 2–3’ jog
4 sets of 3’-2’-1’ at 8k-5k-3k effort w/ 1’ moderate
3 km at 105% HMP, then 2 × (6 × 500 m at 109% HMP w/ 45–60” walk), 4–5’ set rest
5k competition
Athlete C
10 × 2’ cut-down w/ 1.5’ jog
(No other supportive speed!)
Athlete D
3 sets of 3’-2’-1’ at 8k-5k-3k effort w/ 1’ easy
5 sets of 3’, 2’ at 8k, 5k effort w/ 1’ easy
5 × 1 km / 4 × 1 km dual cut-down workout (see note below)
3 sets of 4’-3’-2’-1’ at 10k → 5k effort w/ 1’ moderate
5k competition
4 sets of 3 × 90” at 106–108% HMP w/ 1’ / 4’ jog

Individualizing HM-specific speed (105% HMP)

HM-specific speed is a highly neglected component of training for many half marathoners. When I review historical training data from athletes who come to me for coaching, the two most common "missing pieces" are long fast runs and long repeats at ~105% HMP.

Here's where we put HM-specific speed on the calendar for each athlete:

And here are the key specific speed sessions for each athlete, with some comments on the individualization strategy:

Athlete A: HM-specific speed
4 × 6 min at 8k / 10k effort on XC terrain
6 km at 8k / 10k effort on XC terrain
5 × 2 km at 105% HMP with 3’ jog
10k competition (~32:30, hilly course)

Athlete A was able to build some HM-specific speed in the latter part of the general phase with effort-based cross country workouts. He had a 10k race on the schedule, which is excellent specific speed training.

We actually planned on doing at least one more specific speed workout before the 10k (probably something like 3x3k or 3-2-2-1k) but he had a minor knee injury that interrupted training for about five days. By the time he was ready to work out again it was time for the 10k!

Athlete B: HM-specific speed
8k competition (~27:10, soon after being sick)
1-3-2-1 km at 102 → 105% HMP
5k competition (~15:40)
6 × 500 m at 105% HMP w/ 30” walk, 3’ rest, 3 km at 105% HMP, 3’ rest, 6 × 500 m at 105% HMP w/ 30” walk
16 × 500 m at 105% HMP w/ 30” walk

As a slow-twitch marathoner, Athlete B needed to work for a very long period to build specific speed, since it’s always a weak point for him.

These kinds of runners can’t immediately start with workouts like 4 x 2k or 3 x 3k — they just can’t average a fast enough pace for long repeats early on.

Instead, we used shorter repeats, cut-down workouts, short fast continuous runs, and competitions to put a lot of pressure on improving speed at 105% HMP. Once we’ve finally got his aerobic power in a good place, I knew that he would be in extremely good shape after just a couple long fast runs and HM-specific sessions.

Athlete C: HM-specific speed
4 km at 105% HMP, then 4 × 4’ at 102% HMP w/ 1’ jog
4 × 2 km cut-down from 102% → 105% HMP w/ 3’ jog
2 × (2000 m, 1600 m, 1200 m) at 101% → 105% HMP w/ 3–5’ jog
5 × 1200 m at 105% HMP w/ 2–3’ jog

Athlete C’s strongest event is probably the 10k, so specific speed workouts at 105% HMP are home turf for him. Since I knew he would round into shape very quickly after a few of these sessions, we left them until fairly late in training.

Sessions like 5 x 2k or 3 x 3k at 105% HMP can be very mentally draining, and with Athlete C’s short post-marathon turnaround I didn’t want to spend down too much mental energy on lots of tough workouts at 105% HMP, especially early on training.

With a few more weeks to train, I would have added something like 5 x 2k at 105% HMP or 3-2-2-1k at 103 >> 108% HMP to this progression to round out his HM-specific speed.

Athlete D: HM-specific speed
3-2-2-1 km at 102 → 107% HMP w/ 4’ jog
10k competition (~31:30)
5 × 1 km cut-down from 103 → 110% HMP; 4 × 1 km cut-down from 105 → 110% HMP w/ 2’ jog & 5’ rep/set rest
14 × 500 m at 104% HMP w/ 30” walk
8 × 1 km at 105% HMP w/ 2’

Athlete D did not do a lot of long repeats in college, so his training experience with these types of sessions is fairly limited. We used a similar “start from speed” strategy as we did with his long fast runs, using cut-down workouts and shorter repeats with very short rest to build a foundation now for longer workouts in the future (meaning, in this case, future training cycles). In a year or two he will be doing workouts like 5 x 2k or 3 x 3k at 105% HMP.

The 5 x 1k / 4 x 1k dual cut-down workout was a bit of an experiment: it was intended to be 2 x (5 x 1k) cutting down from ~105% HMP to ~110% HMP within each set. I tested out this session with a few different athletes in the last year, but since then have decided that it does not work very well—the workout did not go very well with most athletes, and I didn’t see a big improvement in fitness afterwards.^[9]

Individualizing half marathon specific workouts

The race specific workouts are the most straightforward, conceptually: we need to run 13.1 mi / 21.1 km at race pace on race day. We need our final biggest HM-specific workout to be such that 13.1 mi at 100% HMP is not a huge leap from what we were able to do in training. Then, we just work backwards from there, trying to “build a bridge” from what the athlete is currently capable of to what they need to be able to do on race day.^[10]

Here is where we used major half marathon specific workouts for each athlete:

And here are the specific workouts, with comments on the individualization strategy I used for each athlete:

Athlete A: HM-specific workouts
9 × (1 km at 100% HMP, 1 km at 85–90% HMP)
5 × 3 km at 97% HMP w/ 3’ walk rest (heat/humidity)
2 sets of (3 km, 2 km, 1 km) at 98–101% HMP w/ 1 km at 85% HMP / 4’ easy

Coming off the marathon, Athlete A retained very good lactate shuttling capabilities, and his first 1k/1k workout was already indicative of strong HM potential.

The 5 x 3k workout did not use float recovery because it was run in very hot conditions: given Athlete A’s strong endurance, I wanted him to come as close as possible to HMP for the repeats to ensure he was ready for half marathon pace (and indeed,  these 3ks were at “100% HMP effort” for the day).

Walking rest made that more feasible in the heat—a workout like 5 x (3k, 1k float) would have been much slower.

The final HMP workout (3-2-1-3-2-1 w/ 1k float/easy) was very tough—after bouncing back from the minor knee issue we were in a bit of a time crunch, and the only good time to do a tough session was three days after a 10k road race.

Normally I do not like doing two very tough sessions in a row, but by that point we only had 12 days before the race. He switched to just jogging the rest on the second set (which was the right choice), but was able to hit the target paces, which gave us a lot of confidence going into race day.

Notice, on the calendar, how we really focus on regeneration and recovery for several days following this 10k race + HMP specific workout block. Big stress, big recovery!

Athlete B: HM-specific workouts
7 × (2 km at 100% current HMP, 1 km at 90% current HMP)
HM race (1:13:00)
2 sets of (3 km, 2 km, 1 km) at 96–98% HMP w/ 1 km at 80–85% HMP (very hot)
AM: 5 mi at 95% HMP / PM: 5 × (1200 m, 400 m) at 101%, 89% HMP
4 × (4 km at 97% HMP, 1 km at 78% HMP)

Athlete B raced a half marathon earlier in his buildup, on relatively little specific training, so we had a good idea of his then-current fitness. His early workouts (like 7 x 2k/1k) were at 100% of current HMP, which we later moved towards goal HMP as his aerobic power improved.

Looking at this progression now, I think I should have added one more workout, or altered the final 4 x 4k session. Adding a workout like 5 x 3k at 100% HMP w/ 3 min walk, or modifying the 4k workout to be 4 x 4k at 98 >> 100% HMP w/ 4 min walk, would have helped him reach goal HMP earlier in training—much like we used 5 x 3k with athlete A to hit HMP in the heat.

Athlete C
7 × 1200 m at 102% HMP w/ 400 m at 85% HMP
6 × 1600 m at 102% HMP w/ 400 m at 88% HMP
3 × 2 mi at 97 → 101% HMP w/ 0.5 mi at 80% HMP

Again, Athlete C was working with very little time to prepare. We did very little HM-specific training, and built up to it “starting from speed” via 1200s and 1600s with float recovery. The 1600/400 workout was actually supposed to be 7 sets, but it was very humid, and these kinds of float-recovery sessions are tough to do in heat and humidity.

We ended up being very pressed for time to do race-specific work, not only because of the short post-marathon turnaround but also because this athlete ended up having very sore legs for nearly a full week following a pickup soccer game. It would have been pointless and discouraging to try to do a HM-specific workout in the middle of that week -- you need to be fresh for big race-specific workouts. So, we waited as long as possible for his legs to freshen up and did one final tune-up session (3 x 2 mi) seven days before the race.

If we had a few more weeks to train, I would have liked to get in a session like 8–10 x (1k at 100% HMP, 1k at 90% HMP) – and you can see how the 1600/400 workout was building towards such a session.

I think with one or two more HMP-specific workouts, one more long fast run (and perhaps skipping the soccer game!), Athlete C might’ve been able to continue this progression and run “102% HMP” for the full half marathon, which would have been under 1:08. Indeed, he did end up running two road races about four months later that both convert to sub-1:08.

Athlete D
8 × (1200 m at 97% HMP, 1 min jog) – probably ~100% current HMP
6 × (1200 m at 99% HMP, 400 m at 88% HMP) – probably ~101% current HMP
8 × (1 km at 101% HMP, 1 km at 92% HMP)
5 × (2 km at 99% HMP, 1 km at 90% HMP)

Athlete D used a similar “start from speed” progression in his HM-specific workouts. Again, as a middle distance runner, he is fairly new to these long workouts with float recovery, so I didn’t want to overdo it.

In fact, the final 5 x (2k, 1k) workout might’ve been a little too tough—he faded a bit late in the workout, which was not ideal for confidence two weeks out from race day.

A longer workout with walking rest, for example 3 x (3k, 2k) at 100–102% HMP with 3 min walk rest might’ve been a better choice. But then again, perhaps a bit of a “crisis” in the body was exactly what he needed to improve!

You don’t want to be afraid of workouts with a higher-than-normal chance of hitting a wall, but ideally you want your last race-specific workout to feel good and have a high chance of success.^[11]

Recap: Elements of success for sub-70 HM training

Training for the half marathon is relatively straightforward: you don’t have to worry about tactics or your balance of speed and endurance like you do in the 800m and 1500m, and you don’t have to worry about fueling or physiological resilience like you do in the marathon.

As the schedules above illustrate, the core parts of modern half marathon training are:

• Building up to a long fast run of 13–15 mi at 95% HMP about 3–4 weeks before the race
• Using Canova-style alternating kilometers with float recovery to run progressively further at 100% HMP
• Specific speedwork consisting of 8–10 km of intervals at 105% HMP and repeats of 1–3 km; possibly also supplemented with continuous runs at 105% HMP and/or competitions of 8k/10k
• Shorter repeats at 107–110% HMP, optionally done as float-recovery fartlek sessions; possibly also competitions of 5k
• High (but not necessarily very high) overall training volume. For sub-1:10, 70–80 mi/wk (110–130 km/wk) is sufficient for a moderately talented athlete.
• A large base of general fitness built using high mileage, high-end aerobic workouts, and effort-based sessions across a wide range of speeds (possibly incorporating hills and/or cross-country terrain to build muscular strength, endurance, and power)
• Training that continuously evolves, instead of repeating the exact same workouts over and over

Finally, remember that individualization matters. These four athletes all run within two minutes of one another in the half marathon, and their training plans follow the same principles, but their actual workouts were often remarkably different.

This is why I think it’s absurd to worry about people “stealing” your workouts. If you just blindly copy what someone else does, without understanding why they do it, you are not likely to get good results!

This percentage-based approach may or may not be "the best" way to train for the half marathon in some cosmic sense, but the schedules above demonstrate why I'm confident that it produces consistent results across very different types of runners in very different training circumstances.

Coaching openings for fall/winter 2025-2026

If you are looking for individualized training to run a fast half marathon this year or next, or a fast performance in other distances, I am currently (as of October 2025) taking on new athletes for individual coaching.

If you think this kind of individualized approach to training would be a good fit for you, reach out to me at john@runningwritings.com and let me know a little bit about your running background and your goals, both for this season and beyond.

Learn more about modern training

Hopefully you’ve found this comparative approach to half marathon training helpful. If you did, you can support my work by sharing this article with friends or posting about it on social media.

If you want a book-length treatment of this percentage-based approach as it applies to marathon training, check out my brand-new book: Marathon Excellence for Everyone. Marathon Excellence is, put simply, your comprehensive guide to modern marathon training and racing.

Indeed, three of the four sub-70 half-marathoners in this article have run well under 2:30 in the marathon (2:24, 2:24, and 2:26 currently) following the Marathon Excellence approach, with the fourth planning on making his marathon debut in the next year.

If you want to find out when I have new articles coming out on advanced training methods, or the latest physiology and biomechanics research, sign up for my newsletter! It’s the best way to find out when I have new content.

Footnotes

---

^[1] Racing a half marathon during a marathon buildup or during a more general road racing season requires a somewhat different approach, using less half marathon specific work and putting the race earlier in the buildup.

^[2] Lest you think these performances are just a matter of working with naturally talented athletes, Athletes A, B, and C all had marathon personal bests slower than 2:40 when they started working with me.

^[3] These volume guidelines are about what I expect for a sub-70 HM runner; slower runners doing less mileage may need to adapt workout volume and intensity depending on their experience, mileage, and ability level. You’ll notice that Athlete D, who has the least amount of road running experience, does less volume in all sessions at 95–105% HMP for this reason.

^[4]  Regarding Athlete C, I would usually not recommend an eight-week marathon-to-half-marathon turnaround, but this athlete bounces back from marathons very well and we did not have any other good racing opportunities for several months anyways. His marathon was very successful and he ran with a negative-split strategy, so I was less worried about post-race recovery.

Renato Canova once wrote that, paradoxically, when one of his athletes runs a 2:04 or 2:05 marathon in a hyper-fast race like Valencia, recovery is very easy—a fast marathon that’s an even-pace or negative-split effort is not that hard on your body if you have good training. Rather, it’s the brutally hot (or brutally cold and rainy) slug-fests on tough courses that really wreck your body and take weeks or months to recover from.

^[5]  For example, we avoid track workouts entirely, opting for time-based fartleks, longer runs on trails or scenic courses, and doing runs and with friends. During the introductory period I’d happily have an athlete do a “suboptimal” workout if it lets them run with a group of friends–this is why Athlete C was doing 6mi fast on the roads less than two weeks after a marathon!

^[6] Renato Canova describes training as “a continuous EMERGENCY” that always needs to adapt to the real situation of the athlete, which I think is a very good mindset to have.

^[7] Throughout Athlete B’s schedule I’m using a mixture of three strategies to plan workouts – in my high-level overview of percentage-based training. I call these “fitness updates”, “workout evolution”, and “bridging to goal pace.” So a workout that, early on in training, might be at 95% current HMP evolves into a different category of workout (at the same absolute pace) later in training.

^[8] For example, during his London buildup, Emile Cairess did 14 x 400m at 109% MP w/ 300m float at 85% MP, followed by 5k progressing from 86% to 110% MP. Eight days later, he did 3 sets of 10 x 400m at 97 / 110 / 112% MP with 1600m moderate between sets. This is clearly not just “VO2max work” like you might be led to believe by someone with a narrow physiologically based view of training, but I don’t have a succinct explanation for why he did those sessions, and why he did them at that point in training (~25–30 days out from race day).

^[9] I think it's important to try new things in training. It's ok if they don’t work out! As Canova says, “Never a good coach finishes to learn.” In the case of the 2 x (5 x 1k) dual cut-down workout, I think the issue is that you can go too fast at the end of the first set, which makes the second set too difficult. Perhaps something like 3 x 2k + 4 x 1k can touch on the 105–110% range more effectively?

^[10] One note regarding race-specific workouts: the races these athletes were preparing for were on flat courses with (mostly) good weather. A more challenging race, like the hilly Philadelphia Half Marathon course for example, would require more preparation in training for the demands of the course. Ditto for a race happening in the heat of the summer.

^[11] With marathoners I will often put the toughest marathon-specific workout 3–4 weeks out from race day, and be okay with them struggling or fading a bit in that session. Then, 10–14 days out from race day, we will do a more controlled session with a high chance of success as a tune-up. See Marathon Excellence for more on marathon tapering.

Related articles