Building a full-spectrum percentage-based 10k training plan from scratch

I’ve written a lot about what I’m now calling full-spectrum training, i.e. percentage-based training in the style of Renato Canova and other members of the “Italian School” of training.

We can talk theory all day, but sometimes it helps to see what building a real schedule looks like from the ground up. To that end, I recently received the following email, lightly edited and anonymized:

I’ve just recently came across your website while looking into Canova-type training and have read a few other topics that you have delved into in great detail!

The last few years I really got into strict Lydiard-type training and while I found this training has really helped me, I think I am at a stage where I need to start looking at different types of training techniques. Hence I came across Canova or percentage type training.

I believe percentage-based training has helped me, but am probably struggling to write a plan for myself. I am looking to aim for a 10 km and want to see if you can give me any guidance on how to structure my training for it. I am keen to get a grip of this myself so I know how to write a plan up for various distances.

I started typing out a response, and quickly realized that it would work better as a full article, since I get a lot of questions like this.

So, here it is: a guide to building a 10k training program from scratch, using the full-spectrum percentage-based approach. This isn’t “the” way to train for the 10k, but it is a way, and one that is in keeping with the modern, full-spectrum, percentage-based approach.

This is also not “the Canova method,” but my method, one that’s certainly been influenced by Canova’s approach, but that also has some important differences and other influences as well.

Full-spectrum percentage-based 10k training schedule (printable PDF)

If you want to see the final schedule right away, or have it open / printed out so you can follow along, here it is:

🗓️ Download the 10k training schedule here

If you prefer the schedule in kilometers (metric) you can download that version here.

As a reminder, I have a race pace percentage calculator that makes it super easy to calculate your paces for percentage-based workouts.

Setting the initial parameters for our 10k training schedule

For the actual schedule, I’ll be assuming we’re working with a sub-elite runner currently doing about 55 mi/wk with some light workouts, with 16 weeks to prepare for a 10k. I’ll also assume that we have a good estimate of our athlete’s current 10k fitness,^[1] and that they’re ready to build up to 70–75 mi/wk over the course of several weeks.

I chose these parameters to give myself a bit of a challenge—it’s easy to say “oh just run 100 mi/wk then add fancy workouts X, Y, Z.” But that’s not very useful for most people.

Instead, I wanted a schedule that would (1) use the full-spectrum training approach, (2) showcase several advanced training techniques, and (3) be workable even with moderate overall mileage. You could push this schedule down to a peak mileage level of 70 mi/wk, and maybe even 65 mi/wk, and still have it be viable.^[2] I also have some thoughts below on how you’d adapt it for even less mileage and a less-advanced approach (i.e. fewer doubles, no double threshold, etc.). 

With these goals, the schedule is probably a little over-engineered: for any particular individual, I might tone down one of these elements (fewer advanced techniques, simpler workouts, higher or lower mileage, etc). But as a coach (and as an athlete) it's good to push yourself a bit sometimes, and this schedule is the result.

Given these starting parameters, the runner’s actual fitness level and gender are not that important; anyone in the 30:00–38:00 10k range would be fine doing this kind of program, and even plus or minus a few minutes on either end of that range is probably okay too.

At a high level, I approach making a training plan with the same philosophy as the training itself. By that, I just mean that I start with the goal: the 10k race itself. That’s the first thing that goes on the schedule.^[3]

Then, I work backwards, dropping in the key race-specific workouts into the schedule. I typically have a progression of race-specific workouts I’d like the athlete to do.

Then, I turn my attention to specific speed and specific endurance, then supportive speed and supportive endurance, and finally general training, including the elements that most people go to first when building a training plan—mileage, easy runs, and long easy runs.

Phases of training: general, race-supportive, and race-specific

I’d like to think about these workouts in the context of the race-specific phase (6 weeks), the race-supportive phase (4 weeks), and the general phase (6 weeks), so let’s sketch those out straight off the bat.

I arrived at the 6/4/6 week structure by aiming for 6 weeks of specific training, then working with however much time we had left. With 20 weeks to train, I’d probably still do 6 weeks of specific training, but increase general training to 8 weeks of general training (and also use 6 weeks of race-supportive training). With even more time, all the rest would go into general training.

For lower-mileage or less-experienced runners, I’d only use 4 weeks of race-specific training, and shift to a longer general phase. As a rule, the more experienced you are, the less general training you need.

The converse is also true: younger, less experienced runners should do shorter periods of race-supportive and race-specific training.

10k-specific workouts

I know that about 10–14 days before the race, we want to do a big 10k-focused workout, but perhaps not the absolute biggest workout of the season.

Likely we’ll do something with a mix of paces at 10k pace plus a bit faster than 10k pace. The specifics of that final session might vary depending on the level of the athlete. Here are a few examples of final 10k-specific sessions:

Athlete level	Mileage	Final session
Mid-level	50–65 mi/wk / 80–105 km/wk	4 × 2k at 100–102% 10k pace w/ 3 min jog
Sub-elite	70–85 mi/wk / 110–135 km/wk	3-2-2-1k at 100, 101, 102, 105–107% 10k w/ 3 min jog
Elite	90+ mi/wk / 145+ km/wk	4-3-2-1k at 98, 100, 102, 105–107% 10k w/ 3 min jog

For this schedule, we’ll settle for 3-2-2-1k as our final session.

About 10–14 days before this final workout, we want another 10k-specific session. This second-to-last (“penultimate”) workout will usually be the longest or highest-volume session. Again, examples would be:

Athlete level	Penultimate 10k-specific session
Mid-level	2-1-2-1-2k at 98–99% (2ks) and 100–101% (1ks) w/ 3–4 min jog
Sub-elite	2 sets of: (3k at 99–100% 10k pace, 3min jog, 2k at 100–102% 10k pace, 4min walk/jog)
Elite	3 × 3k at 98–100% 10k pace w/ 3–4 min jog, then optionally 1k at 105–108% 10k

For this schedule, we’ll use 5 × 2k at 100% 10k as the penultimate session, and put that workout on the schedule as well.

Then we can keep dropping in stepping-stone workouts leading up to that biggest 10k specific session, each about 10–14 days before the last, at least in the final six weeks before the race. Sometimes we can mix these paces in alongside 105% 10k pace (so, about 5k pace), especially in sessions in the race-supportive phase.

Here is the 10k workout progression I settled on for this schedule:

Session	Workout
1	8 × 1000m at 100% 10k w/ 2min jog
2	4 sets of (1200m at 100%, 1min walk, 800m at 103–105%) w/ 4 min walk/jog
3	4 × 2k at 100% 10k pace w/ 3 min jog
4 (biggest)	5 × 2k at 100% 10k pace w/ 3 min jog
6 (final)	3-2-2-1k at 100, 101, 102, 105–107% 10k w/ 3 min jog

You could also use 8k and 10k races as 10k-specific workouts, whether they are on the track, on the roads, or over cross-country.

Sometimes it makes sense to swap in a fast continuous run at 100% 10k pace (e.g. 3–4 mi or 5–6 km continuous) somewhere in this progression, for athletes who do not race very often. Session 3 would be a good occasion for this kind of workout (i.e. 5-6 km continuous, instead of 4 × 2k). These fast continuous runs make less sense if you’re already doing some 5k and 10k races in your buildup, though.

Once we have our workout progression, we can block it out on the schedule:

Notice how the 10k-specific workouts become more common later in training.

10k-specific and 10k-supportive endurance

If race-specific work is most important, what comes next? Race-specific and race supportive endurance, i.e. 95% and 90% of race pace.

To support our 10k-specific workouts, we want appropriately-sized endurance work. Again, some guidelines for the final workout I’d like to see each athlete work up to:

10k-specific endurance (95% 10k)

Athlete level	Mileage	Final session
Mid-level	50–65 mi/wk / 80–105 km/wk	5–6 mi continuous
Sub-elite	70–85 mi/wk / 110–135 km/wk	7–8 mi continuous or 7–8 sets of (1k at 95% 10k, 1k at 80–85% 10k)
Elite	90+ mi/wk / 145+ km/wk	8–9 mi continuous or 5 sets of (3k at 95% 10k, 1k at 85% 10k)

I mostly arrived at these final workouts by asking myself "what kind of work at 95% 10k pace would you need to be able to do, to be ready for the final 10k-specific workout for this athlete?"

It’s reasonable to consider 15k, 20k, and half marathon races as 10k-specific endurance workouts.

Sometimes it also makes sense to blend 90% and 95%, e.g with 12 mi as (4mi at 90%, 4mi at 92%, 4mi at 94–95% 10k). A session like that “counts” as both paces, in terms of the schedule.

I also consider threshold workouts (see below) to be 10k-specific endurance.

10k-supportive endurance (90% 10k)

We can look at a similar final goal session for 90% 10k pace. Many people would call this “marathon pace” but I try to avoid that term unless we are talking about actual marathon-specific workouts. In my first book, I called this pace “aerobic threshold,” though that’s technically a misnomer.

In any case, we want to use strong, continuous runs at this pace. Final workouts might be:

Athlete level	Mileage	Final session
Mid-level	50–65 mi/wk / 80–105 km/wk	8 mi at 90% 10k
Sub-elite	70–85 mi/wk / 110–135 km/wk	10 mi at 90% 10k
Elite	90+ mi/wk / 145+ km/wk	12 mi at 90% 10k

With both the supportive and specific endurance workouts, we can work backwards, building a progression of stepping-stone workouts using shorter runs at the same speed for continuous sessions, or shorter rep lengths for alternating kilometer sessions.

For example, with continuous runs at 95% 10k, we might start at 4mi (not a hard session at all), then add one mile each time we do the workout, hitting that pace every two or three weeks, and eventually reaching 7–8 mi continuous for a sub-elite runner.

With 95% 10k pace, we can also switch between doing continuous runs and alternating kilometers. For example, after doing 5 mi at 95% 10k, maybe in two weeks we can do 6–7 sets of (1k at 95%, 1k at 85%), then two or three weeks after that, doing 6 or 7 mi continuous at 95% 10k, then 8 sets of 1k / 1k the next time we hit this pace. Or do the same number of 1k / 1k reps, but increase the recovery pace. Many different options can work.

These workouts go into the schedule now, with the aim of emphasizing 95% 10k more often as we get closer to the race:

10k-specific and 10k-supportive speed

Moving back to faster-than-10k work, we can follow a similar pattern.

Especially for a longer event like the 10k, speed matters much less than endurance, so we only have to hit these paces every 14–28 days, depending on the phase of training, and we should be more eager to mix them together or use them as secondary sessions (i.e. as a shorter afternoon session following an endurance-focused morning workout).

As above, some guidelines for the final workouts at each pace:

10k-specific speed (105% 10k pace)

Athlete level	Mileage	Final session
Mid-level	50–65 mi/wk / 80–105 km/wk	4 sets of (1000m at 104% 10k, 1 min jog, 500m at 106% 10k, 4min walk/jog)
Sub-elite	70–85 mi/wk / 110–135 km/wk	2 sets of (1200, 1000, 800, 500m) at 102–107% 10k w/ 3 min jog
Elite	90+ mi/wk / 145+ km/wk	4 sets of (1k at 103–104% 10k, 1 min jog, 1k at 105–107% 10k, 3–5min walk/jog)

Doing a 5k race also counts as 10k-specific speed. As earlier, we’d work up to these with shorter, easier workouts in the training plan. I kept it simple on this schedule with 6 × 1k at 105%.

10k-supportive speed (110% 10k pace)

Again, reasonable final workouts might be:

Athlete level	Mileage	Final session
Mid-level	50–65 mi/wk / 80–105 km/wk	2 sets of 4–5 × 500m at 106–109% 10k w/ 55 sec walk between reps and 4–5min walk/jog between sets
Sub-elite	70–85 mi/wk / 110–135 km/wk	3 sets of (600-500-400-300-200m) at 107–112% 10k w/ 2min walk between sets
Elite	90+ mi/wk / 145+ km/wk	3 sets of: 4–5 x 500m at 107–110% 10k w/ 45sec walk between reps and 4–5min walk/jog between sets

For workouts close to 110% of 10k pace, you need to be mindful of whether you are a speed or endurance-oriented runner. For more endurance-oriented runners, 110% 10k pace might be too fast—you may need to cap these supportive speed sessions at 107–108% 10k pace instead.^[4]

You only need to hit 10k-supportive speed (110%) every few weeks. It’s far less important than 90% and 95% of 10k pace, and you could still run quite a good 10k if you completely skipped these workouts if you tossed in some strides and hill sprints.^[5]

If you want to run well in the 5k, or if you’re training for a championship-style 10k race, these workouts take on greater importance—but you should pretty much never neglect supportive or specific endurance to focus on speed.

Here’s where these specific and supportive speed sessions make sense in the overall plan:

Note that on Monday of week 8, we "double count" a workout: it's 4 sets of (1200m at 100% / 800m at 103–105%) w/ 1 min / 3–4 min walk/jog. It's a nice way to get a moderate mixed-speed stimulus at both 100 and 105%.

Some of the supportive speed workouts also reach up to 115%, though exactly how fast you go depends more on the individual athlete.

A high-level view of key 10k workouts

Here’s all the primary workouts in one panel:

Notice how we use Canova’s idea of the “funnel” of specificity, where race-specific work (95–105% 10k pace) becomes more prominent and more frequent later on in training.

It may look like we're "missing" some 100–105% 10k pace work in the general phase, but we'll use more low-key effort-based fartlek to fill in that gap; read on for more on that strategy.

Filling in easy mileage

Haven’t I forgotten the most important thing—mileage? Well, no, not really. Mileage is means to an end, so it only makes sense to think about mileage after we’ve gotten the key workouts squared away.

The demands of the event put constraints on what kind of workouts we need to do, so we need to think about what kind of easy run distance we need to support our 10k-specific workouts. Then we can figure out how much mileage we can fit into a week, informed also by injury risk, training history, and long-term running career plans.

In the case of the 10k, even a relative novice needs to be ready to do at a minimum a few workouts with 8 km of quality at 10k pace.

Even with a pretty short warm-up and cool-down, you’re probably looking at an 8 mile day for that kind of workout. After considering the requirements for doing 8–10 mi strong, and 10–12 mi steady, the requirements for easy mileage become more clear.

For a mid-level runner, for example, before moving to 10k-specific and 10k-supportive workouts, we need to be fit enough to do:^[6]

• 7–8 miles easy on a regular basis (at least a few times per week)
• 9–12 miles easy fairly often (once every 5–12 days)
• 12–15 miles easy once in a while (once every 2–3 weeks)

Once we know what kind of individual easy runs we need, we can start figuring out a mileage progression that makes sense. We can use easy runs, moderate runs, and doubles to fill in mileage as needed.

It’s fine for total mileage to drop during the race-specific phase, since the focus needs to be on recovering well after big training sessions.

So, we can add mileage to the training plan now, keeping in mind that progression in mileage should be in service of the progression in workouts—not the other way around.^[7]

Here's what mileage looks like (in miles and in kilometers):

General endurance and supportive workouts

The remainder of the schedule can be filled in with general fitness work and secondary workouts. Many different workouts can fit in well here, but there are a few sessions I’d like to specifically highlight.

Fartlek sessions

As noted above, right now it looks like we have some "gaps" where we don't touch on 100% or 105% 10k pace very much in the first several weeks of training.

Doing some light fartlek sessions, like 8–10 × 2 min at 105% or sets of 3-2-1 min at 100–108% 10k is a good way to touch on those speeds in general training in a low-key fashion.

Long fast runs at 85% 10k

These sessions are useful every 3–4 weeks throughout training, and help serve as endurance support for the 10k supportive endurance work.

If, for example, we want a sub-elite runner able to do 10 mi at 90% 10k pace three weeks out from the race, that runner should be able to do 12–13 mi at 85% 10k pace 3–4 weeks prior to that 90% workout, and on downward. Progressing these runs is straightforward; you just go further at the same speed.

Long easy and long moderate runs

You can do some long easy (~55–65% 10k pace) and/or long moderate (~65–75% 10k pace) runs during general training, but these really only need to be long enough to support the long fast runs.

So, you can easily get away with going no further than 15 miles (which is what I’ve chosen for this schedule).

Doing up to 18 mi is fine, but instead of progressing the distance beyond 15–16 miles, for the 10k it’s usually better to move to long fast runs at 85 and 90% 10k pace, abandoning long easy completely once you’re ready to go long and fast.

Kenyan-style progression runs

These runs start extremely easy, much slower than a normal run, and very gradually work their way down to a strong aerobic pace through the middle. By the last few miles you might be cruising along around half marathon effort, and perhaps close a bit faster than that in the final 1–2 miles. The exact splits are less important; effort is the focus.

I would normally use a few more of these, but in trying to showcase the more advanced workout types, they tended to push out Kenyan-style progression runs, so there are only three on the schedule. But you could do a lot worse than just replacing a lot of the more advanced workouts with progression runs.

I have a more detailed article in the works on how to do Kenyan-style progression runs, so make sure you join my email list to find out when that comes out.

Moderate runs

Once or twice a week, especially in the general and race-supportive phase, it is a good idea to consciously do a run at more of a moderate effort (~70–80% 10k). Doubly so if you’re more experienced, in which case you can also do occasional shorter runs at a steady to strong effort (~85–90% 10k).

Here's how these secondary workouts fit into the schedule:

Moderate runs taper off in the race-specific phase, since we're focusing more intently on being fresh for big workouts.

Advanced workout types

One of the things I tried to showcase in this schedule is how to incorporate a range of different advanced workout types. By advanced I mean modern, innovative workouts that are relatively recent developments, and are not commonly used by amateur runners or runners doing explicitly “old-school” training.

Float recovery for traditional lactate threshold workouts

One way of viewing traditional threshold workouts (e.g. Jack Daniels-style cruise intervals) is as 10k-supportive endurance, since lactate threshold pace^[8] tends to be ~96–97% 10k pace for runners under about 40 min in the 10k.

I prefer to see traditional threshold work as a precursor to 10k-supportive endurance, so my approach is to use traditional threshold work fairly often in general training, then mostly phase it out in favor of continuous workouts at 90% and 95% 10k pace—either in the form of fast continuous runs or alternating kilometer sessions.

There’s an easy way to bridge from threshold work to alternating kilometers, via the use of float recovery during cruise interval sessions. Here’s a typical progression I’ve used successfully with many athletes, adjusted here to be appropriate for a mid-level (50–65 mi/wk):

Session	Workout
1	6 × 1200m at 96–97% 10k pace (~Daniels “T pace”) w/ 1 min jog rest
2	6 × 1200m at 96–97% 10k w/ 400m float at 80% 10k
3	6 × 1200m at 96–97% 10k w/ 400m float at 85% 10k
4	6–7 sets of (1k at 95% 10k, 1k at 80–85% 10k)

Th real key to improving with float recovery workouts is increasing the speed of the float recovery over time (and from one season to the next). This requires precise pacing of course but the rewards are enormous. A very fit runner could do a session like 6 × 1600m at 96–97% 10k, recovering with 400m as fast as 90% 10k pace. Of course, you don't start with that; you get there over the course of several years.

Double threshold

All of the above applies to Norwegian-style double threshold sessions as well, though it’s worth thinking about how the so-called sub-threshold sessions and supra-threshold sessions fit into the full-spectrum percentage-based system.

I’ve been experimenting with double threshold workouts more, for runners of all levels, with reasonably good results so far, but I’m still on the fence as to how I feel overall about them. In the past, I would use more Kenyan-style progression runs and longer runs of 11–15 mi with a progressive finish or sub-threshold repeats at the end (e.g. 3 × 1.5 mi at 91–93%). I still think those classic high-end aerobic workouts are a safe bet if you don't want to do double threshold.

Sub-threshold sessions

A classic Ingebrigtsen-style session is 4–5 × 2000m at a speed backed off a bit from classical lactate threshold (LT2), with 1–2 minutes of walk recovery. Instead of 96–97% 10k pace for classical threshold work, these “sub-threshold” sessions might be more like 91–93% 10k pace.^[9]

These sub-threshold sessions can be used as a foundation for 10k-supportive endurance, and your eventual goal in training can be to move towards doing fast continuous runs at this same pace. In other words, 91–93% 10k pace can, over time, become 90% 10k pace.

Supra-threshold sessions

Another common Ingebrigtsen-style double threshold session might be a “supra-threshold” session like 16 × 400m at a bit faster than classical lactate threshold pace, with only 20–30 seconds of walking rest. It’s quite possible to run these basically at 100% 10k pace without accumulating lactate from one rep to the next.^[10]

For that reason, these workouts can serve as a foundation for later 10k-specific work. You can even bridge your way towards true 10k-specific work with a progression like this, doing one session done every 10–14 days (though this schedule does not use this technique).

Session	Workout
1	12 × 500m at 99–101% 10k w/ 25 sec walk rest (as evening double threshold session)
2	3 sets of (500m, 500m, 1000m) at 99–101% 10k pace w/ 25 sec walk within sets and 2-3 min walk/jog between sets
3	4 × 500m at 99–101% 10k w/ 25 sec walk rest, then 2 min walk/jog, then 3000-4000m at 98–100% 10k pace

Evening speed sessions

These evening speed sessions aren’t double threshold at all, but I still think of them as advanced sessions within the broader family of double threshold-inspired workouts. Consider the following training day:

AM: 4 × 2000m at 91–94% 10k pace w/ 2min walk/jog rest

PM: 2000m at 96% 10k, 3–4min walk rest, then 2 sets of 600-400-300-200m at 108–115% 10k pace w/ 2–3min walk rest and 4–5 min walk/jog between sets

This might look more like a Renato Canova-style special block, though it eschews the super-high workout volume you’d see in a traditional special block and instead takes inspiration from double threshold workouts by having both the AM and PM sessions be restricted to only about 60–75% of a “full” workout.

These evening speed sessions help solve a problem with 10k-supportive speed—where do you put it? It’s inconvenient to dedicate an entire day to ~110% of 10k pace, but you don’t want to completely neglect it. Doing a moderate volume of speed as a secondary session lets you still get some supportive endurance via sub-threshold work in the morning.

Sessions like this are a good option every two to three weeks in the general phase, and perhaps the race-supportive phase as well. These are the most experimental sessions of the advanced workout types—I’ve only used them with a couple of athletes, albeit with good results so far.

(Also, unlike double threshold and special blocks, I didn't "borrow" these ideas from anyone else, I just made it up myself! So, buyer beware!)

Special blocks

This schedule incorporates one special block, about four and a half weeks out from the race. The session is:

AM: 4mi at 85–90%, 3–5 min rest, 4 km at 102–103%

PM: 4mi at 85–90%, 3–5 min rest, 2 × (600m, 400m, 300m, 200m) at 110–115% w/ 1 min walk & 4–5 min walk/jog

In Canova’s system this is a speed-focused special block^[11] that focuses on fast continuous running in the morning and supportive speed in the evening. Even though this special block does not quite reach the heroic volumes you’d see in elite-level Canova-style training, the schedule still uses a significant reduction in volume before and after to recover, since it's a very tough day nonetheless.

Here are the various advanced sessions on the calendar:

When (and whether) to use mixed-speed sessions

You may have noticed that many of the sessions that were nominally at 105% or 110% 10k pace actually involve a range of different speeds, e.g. cut-down repeats from 110–115% 10k pace.

I find these mixed-speeds sessions most useful for race-supportive and race-specific speed workouts. My general rule is that I only mix adjacent “rungs” on the ‘ladder of support’—i.e. my feeling is that mixed speed workouts should only span a range of about 5% of race pace.

This isn’t a strict rule (the 600-400-300-200m evening speed workout above bends it a bit), but it’s a good guideline to keep you from overcomplicating things.

I often see people trying to galaxy-brain their training with intricate workouts like The Michigan^[12] (1600-1200-800-400m fast on the track, with 1.5 mi road tempo in between each track rep), or advanced double ladders like 800-200 / 600-300 / 400-400 / 300-600 / 200-800m.

However, unless you really know what you’re doing, these workouts just end up being a little bit of everything, with not enough stimulus in any one direction to actually improve.

I very rarely use speeds that span more than about a 5% pace range in the same workout. The exceptions are things like fast, relaxed 200s at the end of some threshold work, or occasionally “broken reps” like (600m at 3k pace, short rest, 200m fast)—and even those sessions are mostly for high schoolers who need to be ready for championship-style races.

Now that most of my athletes are post-collegiates, we basically never have to worry about sit-and-kick situations.

I do see professionals use mixed-speed workouts fairly often—Renato Canova often uses these sorts of double ladder sessions; Ron Warhurst (who coached Hobbs Kessler, Kevin Sullivan, Nick Willis, and Alan Webb, among others) invented the Michigan, obviously; and Alberto Salazar (yes, I know)^[13] often had Oregon Project athletes do pyramids (e.g. 2-3-4-5-6-5-4-3-200m), down-ladders, broken reps like (600m, short rest, 200m), and mixtures of fast track repeats and uphill repeats.

I am humble enough to admit that I don’t understand the rationale for most of these advanced workouts, and for that reason I do not use them, outside of sporadic experimentation (sometimes on myself). If you have a strong case or rationale for these kinds of workouts, let me know in the comments!

Intervals versus continuous runs for 10k training

One last point worth mentioning is the balance between interval workouts and fast continuous sessions / "tempo runs." I only bring this up because some of the more adamant defenders of Norwegian-style training insist that intervals are strictly superior to uninterrupted fast continuous runs.

For my part, I believe that intervals and fast continuous runs have complementary strengths and weaknesses—both have a place in training, and both are used often in this schedule:

Racing more often, and replacing workouts with races

I don’t like schedules that presume you’ll be able to find a good race to run on an arbitrary day. Usually you have to work backwards from what races are available in your area.

At least in the United States, it is surprisingly hard to find competitive 10ks, even for runners in the 34–38:00 range, to say nothing of the 33–30:00 crowd. So, usually you have to find your target races, then work with what you’ve got. So, that’s why I haven’t scheduled any races aside from the target race.

In this schedule, it would be pretty reasonable to race a low-key 5k, 8k, or 10k once or twice during the general phase, and a 5k, 10k, or 15k/10mi/20k or half marathon at some point in the race-supportive phase. Week 14 could also be a good opportunity for a 5k, 8k, or 10k. If you race, you can consider the race to be a workout in the appropriate category (e.g. specific speed if you race a 5k, since 5k pace is about 105% 10k pace).

After the 10k in week 16, you can easily have a “racing season” that lasts around three to five more weeks. You could do a 5k as soon as one week later, though 10–14 days would be preferable. You can race another 10k, or even a half marathon, two weeks later.

Beyond three to five weeks, you start needing a more exotic multi-peak schedule, which is a topic for another day and in any case is something only college and professional runners need to worry about.^[14]

Adapting the schedule for lower mileage

If you wanted to adapt this schedule to lower mileage, or a less-advanced 10k runner, what would you change? Here’s what I would start with:

• Turn most or all moderate runs into easy runs.
• Drop mileage for all weeks down by 10–20%, reducing doubles or eliminating them entirely. This would bring the peak mileage to as little as 55–60 mi/wk.
• Eliminate evening speed sessions and just do the morning workout. Compensate by doing 5–6 x 100m strides or 5–6 x 10 second hill sprints twice per week.
• Eliminate double threshold workouts. Alternate between doing the morning sub-threshold session and the evening supra-threshold session, or replace with a Kenyan-style progression run.^[15]
• Reduce the volume of workouts at 85–95% 10k pace by about 10%, and likewise for workouts at 105–110%. Even for a low-mileage runner I would not do less than 8 km of work at 100% 10k pace for the final workouts during the race-specific phase, since the 10k is ten kilometers long for everyone. 

Theoretical schedules are easy; real training is hard

My coaching mentor Jamie Kirkpatrick, who was (and still is) the head coach at Edina High School when I was an assistant coach, often said that the difficult part about coaching isn’t writing the schedule—it’s knowing how to adapt it. He would carry around print-outs of the entire season plan and give them to other coaches if they asked about our training. “It’s not about the schedule, it’s about how you change it,” he’d say.

In contrast, writing this schedule—for a theoretical 10k runner who never gets hurt, never gets sick, and never has any scheduling constraints—is easy.

Too easy, in fact: it gives the illusion that coaching a runner to a 10k PR is something you can sit down and do in an afternoon (as I did when writing this article). In reality, the schedule must change and evolve on a constant basis as the situation changes.

When the situation changes—because of illness, travel, or even workouts that go better than expected—the training plan needs to change too. Sometimes you need to rebuild the plan from scratch multiple times during a training cycle!

Realistically, I’d only expect an athlete to actually complete about 80% of these workouts in a real training cycle—even if training were going very well. Things come up and workouts need to get shifted, rescheduled, modified, or eliminated.

Another thing that made this theoretical schedule much easier than a real schedule is that I got to set the starting parameters: I assumed our 10k runner is healthy, ready to train, and able to start immediately at 55 mi/wk with some light workouts. Moreover, I assumed we had a good estimate of their current 10k fitness.

In the real world you usually have to “meet in the middle,” building a bridge from where the athlete is currently to the key race-specific workouts. Sometimes you need a lot of creativity and ingenuity to get from here to there!

So, the final step of planning training is to continually adapt and revise the schedule, according to the real situation you’re dealing with at the moment.

Final product: 16 weeks of full-spectrum training for a sub-elite 10k runner

The full 16-week plan is downloadable as a PDF at the links below:

🗓️ Download the 10k schedule here (volume and runs in miles)

🗓️ Download the metric 10k schedule here (volume and runs in kilometers)

You’ll want to make use of my free race pace percentage calculator when you plan out your workouts.

Keep in mind this is still just an example schedule, tuned for a sub-elite runner capable of building up to ~75 mi/wk. You should not just blindly copy this schedule yourself without individualizing or adjusting it. I never prescribe exactly the same training schedule to different people, even to two athletes with the same PRs and the same goal race!

Some practical aspects of following the schedule

Initial fitness and revising your estimate

This plan assumes you are starting with a reasonable estimate of your current 10k pace.

Every four weeks or so, you should review your training and update your fitness estimate if necessary. There’s a bit of an art to this revision of your fitness estimate, but as your fitness improves you’ll see your workout paces progressing across the board. As you become more confident in your growing fitness, you can update your 10k fitness estimate if the workouts are getting too easy.

You want to rely most strongly on your paces in workouts at 90–105% 10k pace as a guide for updating your fitness: it’s harder to “cheat” on these workouts, and they're most relevant for predicting 10k performance.

Easy run distances and mileage goals

You’ll need to adjust easy run distances and frequency / distance of your doubles to hit the mileage goals, which is why most easy runs are prescribed as a range of distances. Your individual warm-up and cool-down mileage will also affect the weekly total, so plan (or adjust) your easy run distances accordingly.

Easy run pace guidelines

For non-workout days, this schedule prescribes runs at very easy, easy, moderate, and strong paces. These are primarily effort levels, but if you need pacing guidelines, the following approximate percentages are a good place to start.

Effort level	Approximate pace
Very easy	45–55% 10k
Easy	55–65% 10k
Moderate	70–80% 10k
Strong	85–90% 10k

In general, people obsess too much about getting easy run paces right. With percentage-based training, it’s not that important anyways, since the whole point is to touch on a broad range of speeds. When in doubt, err on the easier side of easy.

Doubles vs. singles

I usually find doubles to be a less-stressful way to hit the same total mileage, but if your schedule is not conducive to doubling you can distribute the easy doubles into the rest of the week’s mileage by doing longer single runs.

Putting it all together: a complete, modern 10k training program

In sum, our strategy for building a full-spectrum percentage-based training plan for the 10k from scratch has been:

1. Analyze the specific demands of the event (run 10 km as fast as possible)
2. Identify the key 10k-specific workouts necessary to perform well in the event (long intervals at 10k pace)
3. Plan out a gradual progression of 10k-specific workouts
4. Identify the key specific endurance (95% 10k pace) and supportive endurance (90% 10k pace) workouts that support your ability to do 10k-specific workouts
5. Plan out gradual progressions for these specific and supportive endurance workouts.
6. Identify the key specific speed (105% 10k pace) and supportive speed (110% 10k pace) workouts that support your ability to do 10k-specific workouts
7. Plan out gradual progressions for these specific speed and supportive speed workouts
8. Identify the easy mileage and general endurance workouts (fartlek, long easy run, etc.) necessary to support all key workouts
9. Plan out a gradual progression of mileage, easy runs, long easy runs, and general endurance, adapting mileage to the requirements of the workouts—not the other way around
10. Continually adapt and revise the schedule as the situation changes

There’s one other aspect that doesn’t neatly fit into a ten-step procedure, which is that when you try to fit everything into the schedule, you inevitably find that it won’t all fit.

So, you have to do a bit of iterative refinement on what workouts go where, how often you hit certain paces, and so on. Finding the right balance between adding extra workouts and putting in some breathing room to recover and have less-structured sessions is a natural part of planning schedules.

Some features of the training plan

In closing, a few distinctive features of this training plan are worth pointing out:

• The schedule uses workouts early and often, but these initial workouts are very light.
• The plan gradually increases workout recovery as training becomes more advanced, moving from only one easy day between workouts to up to three days before the next session.
• Workouts include flexibility, allowing you to feel out the exact pace or number of reps according to how you feel.
• Mileage is less important during the race-specific phase.
• Likewise, the pace of easy and moderate runs is also not that important, especially in the race-supportive and race-specific phases.
• All workouts develop over time, either in terms of their volume, extension, or recovery
• The schedule uses several modern training elements, including double threshold, float recovery, and long fast runs.

The comments section below is open for questions and criticism; fire away.

Learn more about modern training

Hopefully you’ve found this from-scratch approach helpful. If you did, you can support my work by sharing this article with friends or posting about it on social media.

If you're training for the marathon (or hope to in the future), I have a new book out, Marathon Excellence for Everyone, which shows how to apply the same percentage-based approach to the marathon.

Learn more about Marathon Excellence here, or get the book on Amazon.

If you live outside of the United States, Marathon Excellence is also available in a Metric Edition with all workouts in kilometers!

You may also enjoy my enjoy my 2013 book, Modern Training & Physiology, which provides a simpler view of scientifically based training for the 800m to the 10k.

Footnotes

---

^[1] To get an initial fitness estimate, the easiest way is to use a recent race. Perhaps our athlete ran a 10k in the last several weeks, or perhaps we have a converted 5k or half marathon time. If not, you can use your best guess. The initial fitness estimate is not that important, since you can revise it quickly if it’s not accurate.

^[2] Doing this schedule with peak mileage of only 65 mi/wk would be totally viable if you’ve run 75–85 mi/wk in the semi-recent past, e.g. as a part of a previous marathon buildup. The main reason to do higher mileage for the 10k is to support your ability to do bigger workouts; once you gain that ability, you can do workouts that are on par with someone doing, say, 75–85 mi/wk, but drop your easy volume down so you only actually total 65 mi/wk.

^[3] For reasons explained later (here), I don’t assume any other races from the get-go; those tend to be hard to schedule a priori, as they depend a lot on what’s available in your area.

^[4] For speed-oriented 10k runners, the way to approach endurance work is to do lower volume initially, and increase the workout volume more gradually. Given the outsized importance of endurance for the 10k, it’s important that we not compromise too much on the pace when it comes to endurance work—better to do shorter reps, or shorter run distances, vs. backing off the speed too much.

^[5] I did not put strides or hill sprints on this schedule, partly because this article was getting long enough already and partly because it’s trivial to add them to an easy day. Generally they work best on the second easy day after a workout.

^[6] Other workouts introduce their own constraints as well; e.g. if you’re going to be doing double threshold work, you’d better be pretty comfortable doing easy doubles beforehand.

^[7] You actually do not need to maintain your weekly mileage level to maintain the aerobic base you gained from doing easy mileage. You just need to maintain the ability to do the individual run distances. Concretely, let’s say you build up to 70 mi/wk, using a mix of 8–10 mile easy runs and 7 mi + 5 mi doubles. All you need to do to maintain that fitness is ensure that you’re able to occasionally (as in, 2–4 days a week) do an 8–10 mi easy run, and occasionally do a 7 mi + 5 mi double. If the other non-workout days are shorter, it’s totally fine: you’re not going to lose any fitness. With long easy runs, you can completely replace them with long fast runs: 10 mi at 90% 10k pace is a much more potent and specific stimulus than 14 miles at 60% 10k pace. If you have marathon ambitions in the near-ish future it can sometimes make sense to leave in the occasional long easy run, but it won’t be helping your 10k fitness much.

^[8] Meaning LT2 pace, not LT1.

^[9] A typical goal for sub-threshold workouts in the Norwegian system might be averaging 2.5–3.0 mM lactate, as opposed to the classical 4 mM of lactate associated with LT2. See my article on SSmax for more on the issues with targeting a certain lactate number—the important part is stability over time, not the exact number.

^[10]  I have a very old article about using faster-than-threshold paces as supra-threshold sessions; though today I only agree with about 80% of what I wrote back then, you still might find it an enlightening read.

^[11] Or an “intensive-extensive” special block, in the terminology used in Canova’s marathon training book.

^[12] The unusual structure of The Michigan workout is partly a consequence of the location: Michigan’s track had easy access to a 1.5 mile road loop for tempo work. In later years, I have heard that the team switched to 0.8 or 1.2 mile tempo segments between reps, simply because the original stadium loop wasn’t runnable anymore.

^[13] Someday I’ll tell my Alberto Salazar story. But not today.

^[14] High schoolers too, but of course they don’t race the 10k!

^[15] Usually if someone is not going to do double threshold I recommend doing just the morning session (or evening session) and increasing the volume 20%. But since a lower-mileage version is already reducing overall volume by ~20%, you can keep the workout volume as-is for the individual sessions.

Related articles