Steady-state max (SSmax) for runners: The maximal metabolic steady state

Steady-state max is the single most important concept in exercise physiology. That’s a bold statement, but I’ve come to believe it.

The term “steady-state max” and its abbreviation “SSmax” (just pronounced “steady-state max”) refers to the highest metabolic power output that your body can achieve while maintaining a metabolic equilibrium—i.e. an internal state in your muscles and bloodstream that’s stable over time.

SSmax is connected to topics you’re probably familiar with, like lactate threshold (LT2, technically speaking) and critical speed, but they are not quite the same thing—metrics like LT2 and critical speed are ways of estimating SSmax, which is the actual underlying biological capability.

Understanding steady-state max is absolutely essential for building a modern, physiologically-based approach to training, so let’s dive in.

Origin of the term “steady-state max”

I made it up myself. No, seriously, I did. Physiologists call SSmax the maximal metabolic steady-state, or MMSS, neither of which roll off the tongue particularly well, so we’re going with “steady-state max,” or SSmax for short. The abbreviation parallels nicely with “VO2max” and “HRmax,”^[1] and spares us the problem of trying to remember to say “maximal” instead of “maximum.” So, SSmax it is.

“SSmax” keeps the focus where it belongs: on the fact that it is the highest possible steady-state your body can achieve. The term “SSmax” doesn't emphasize any one specific metric, like oxygen or lactate, helps remind us that we’re concerned with the overall state of your body across time, not treating any one metric as the sole indicator of your body’s status.

Someday, an enterprising researcher might come along with a new blood test for reactive oxygen species, catecholamines, or some other exotic indicator that outperforms lactate, oxygen consumption, or intramuscular phosphocreatine levels when it comes to detecting or describing the physical state of your body during exercise. If and when that happens, the concept of SSmax will remain, even if a metric-based marker like lactate threshold is swept into the dustbin.

What is a metabolic steady-state?

Let’s start by considering what happens inside your body during a 30-minute easy run. For simplicity’s sake, we’ll assume you’re running on a treadmill in a cool, well-ventilated room. We’ll look at a few key metrics of the overall metabolic state of your body:

• Oxygen consumption, or VO2, which is the rate (in liters per minute, or L/min) that you absorb oxygen from the air.
• Heart rate, which is how quickly your heart is beating (in beats per minute, or bpm).
• Blood lactate, which is the concentration of lactate molecules in your bloodstream (in millimoles of lactate per liter of blood, or mM).

We could also look at a few other metrics, though they’re more difficult to measure during running:

• Intramuscular acidity, or pH (which is also its unit), which measures the concentration of hydrogen ions inside your muscles.
• Intramuscular phosphocreatine, or PCr, which measures your reliance on the anaerobic phosphocreatine system to produce energy.

These variables give an overall picture of the state of your body—collectively, they capture both the function of your heart and lungs (VO2 and heart rate) as well as local conditions inside your muscles (blood lactate,  intramuscular pH, and intramuscular PCr).

During an easy run, all of these metrics behave in a qualitatively similar way: for the first few minutes of the run, they may change in some direction, but after a few minutes, all of these metrics are stable over time. This is the definition of a metabolic steady-state.

Here are some real data—well, real-ish; the plots below are cobbled together from a few different sources—showing what these key metrics look like at an easy pace:

Data here and below adapted from Zuccarelli 2018, Beaulieu 1995, and Heck 1985.

Again, the emphasis here is stability over time: even if you were to come back 15 or 20 minutes later, these variables would still be hovering around the same value.

(You might notice that heart rate reaches a steady-state more slowly; that has to do with heart rate kinetics and cardiac drift, which is a topic for another day!)

Two of these variables are worth a closer examination: VO2 and lactate.

VO2 measures your aerobic system’s metabolic power output

VO2 measures your rate of oxygen consumption, so it’s a direct reflection of the metabolic power output of your aerobic energy system. “Oxygen cost” and “aerobic power output” are two sides of the same coin: running costs energy, and your aerobic system produces that energy—at speeds below SSmax; we’ll get to what happens at faster speeds in a moment.^[2]

An increase in oxygen consumption necessarily reflects an increase in the metabolic cost of running, and therefore, an increase in your metabolic power output. Because oxygen cost is stable over time—and because we’re running on a treadmill at a constant speed—the metabolic cost of running isn’t changing over time.

It is VO2, not heart rate, that is the best reflection of your aerobic system’s metabolic power output. Heart rate can be affected by things like changes in blood plasma volume, but VO2 is strictly tied to the oxygen flux inside your body: a stable VO2 means stable metabolic power output.

Blood lactate measures the balance of lactate production and lactate reuptake

Blood lactate is a concentration, in the sense of being the literal number of lactate molecules per unit volume of blood. Like oxygen, lactate is constantly entering and leaving your bloodstream—even when running at an easy pace—so the blood lactate we measure at any given time is going to be a reflection of the balance between lactate production and lactate reuptake.

Lactate production happens inside muscle fibers. When glucose or glycogen are broken down for energy, the immediate product is pyruvate, which is quickly converted to lactate. Some of this lactate gets shuttled into the mitochondria to get burned for energy using oxygen, but some of it is shuttled out of the muscle fiber. The faster you run, the more lactate you produce—even at easy to moderate paces.

This point is worth repeating: lactate production inside muscle fibers increases as a function of speed, always!

Lactate reuptake is what happens after lactate leaves a muscle fiber. Often, a lactate molecule produced in one muscle fiber is shuttled immediately into another adjacent fiber—a common pattern is that a fast-twitch fiber will split glycogen for energy, then shuttle the resulting lactate next door to its slow-twitch neighbors to offload the work of aerobic respiration (since slow-twitch fibers are better-equipped to oxidize carbohydrates, of which lactate is one).

Lactate can also get shuttled into the bloodstream, then circulate around the body to get taken up by muscle fibers elsewhere in the body. Primary “sinks” of lactate from the blood are other working muscles, the heart (which is the most important working muscle of all), and the liver (which converts it to glycogen) and the kidneys (which convert it to glucose). Like lactate production, lactate reuptake also increases as a function of speed.

So, the blood lactate level that you observe at any given time reflects the balance between lactate production and lactate reuptake.^[3] The stability of blood lactate levels over time reflects whether or not lactate production and lactate reuptake are in equilibrium. In the case of easy running, where blood lactate levels are not significantly different from their baseline values after an easy warmup, these two variables clearly are in equilibrium.

Key physiological metrics at a metabolically unstable speed

In contrast, let’s say we cranked up the treadmill to your 5k pace.^[4] That would produce a decidedly unstable metabolic state, as the following plots show:

Note the contrasts: VO2, heart rate, and lactate are all “crashing”—heading out of control towards some terminal, limiting value. At no point during this continuous run at 5k pace are any of these metrics stable and unchanging.

Some of these limits are familiar: heart rate rapidly approaches HRmax, VO2 rapidly approaches VO2max, and blood lactate, pH, and PCr levels approach, well, whatever maximum (or minimum) level your body can tolerate.

Let’s zoom in again on VO2 and lactate to see what they reveal about the body in a metabolically unstable state.

VO2 at a metabolically unsustainable speed

The plots above clearly show oxygen consumption (VO2) diverging when running at 5k pace. VO2 steadily rises until it hits VO2max, and soon after—usually within a couple of minutes—you’re so fatigued you have to stop running.

The fact that VO2 is not stable at this speed reveals something important: the effective metabolic cost of running is increasing over time. Over the first several minutes of the run, muscle fibers are progressively losing efficiency, meaning they produce less force for a given amount of oxygen.

To avoid slowing down, your body recruits more muscle fibers to pitch in, but that increases the oxygen cost of running. These newly-recruited muscle fibers are more likely to be fast-twitch fibers, which are less efficient and less fatigue-resistant. Soon, these newly-recruited fibers start becoming fatigued and losing efficiency too, causing a cascading effect that drives up the oxygen cost of running.

Blood lactate at a metabolically unsustainable speed

Blood lactate, too, shows a rapidly diverging pattern at 5k pace: it never stabilizes, and instead spirals upwards until you become too fatigued to continue.

Recall from earlier that blood lactate concentration reflects the balance between lactate production and lactate reuptake. Here, lactate production really is outpacing lactate reuptake, for two reasons.

First, fast-twitch fibers—recruited in part because of the loss of muscular efficiency mentioned above—rely more heavily on glycolysis (or the “anaerobic system”) for energy, so the greater the degree of fast-twitch fiber recruitment, the greater the production of lactate.

Second, as your body’s aerobic energy production capabilities top out, your anaerobic energy production continues to ramp up to meet the rising metabolic demand, again leading to greater rates of lactate production.

So, the steady rise in blood lactate at metabolically unsustainable speeds is evidence of both a progressively greater reliance on fast-twitch fibers, and the steady depletion of your anaerobic energy reserves.

As predicted by most models of anaerobic energy production, you can continue at a given speed only for as long as your anaerobic energy reserves last. Once these are depleted, you’re toast—you either must stop or slow down significantly.

Lactate is not the full story when it comes to metabolic steady-states

Lactate and VO2 function as a particularly useful window into the internal state of the body, but SSmax is about more than just one physiological metric.

Metabolic stability is about the ensemble of physiological systems that show stability over time: a stable VO2 reflecting stable metabolic cost of running, a stable intramuscular PCr indicating no progressive depletion of anaerobic energy reserves, a stable intramuscular pH indicating no deterioration in the acid-base balance inside your muscles, and—usually^[5]—a stable blood lactate level reflecting a balance between lactate production and lactate reuptake.

Research using repeated leg extension exercises (a situation more amenable to real-time monitoring of PCr and pH) show the same steady-state dynamic when using critical torque—analogous to critical speed—as the estimate for SSmax:

Figure from Jones 2008

Moreover, a metabolic steady-state is not characterized by any “magic number” like a blood lactate level of 4.0 mM). Again, it is stability over time that defines a steady-state.

If you had to pick just one metric to represent the metabolic state of the body, VO2—oxygen consumption—is your best bet. Since oxygen is the coin of the realm for the aerobic system, stable oxygen consumption over time is strong evidence of metabolic stability.

SSmax separates stable and unstable metabolic power outputs

We saw that running at an easy pace produces a metabolic steady-state, and a very hard pace produces a metabolically unstable state. What happens if we choose an intermediate speed?

At a moderate pace—say, 60% of 5k pace—you’ll get a metabolically stable situation that resembles that seen in the easy run, just with some of the metrics (like VO2 and heart rate) stable at a higher value.

Likewise, at a still-quite-fast speed—say, 6k pace—you’ll still be in a metabolically unstable state, but you’ll arrive at your maximal / limiting values for VO2, heart rate, lactate, etc., a little more slowly.

You can see where this is going: there will be a speed, somewhere in between “moderate” and “all-out 6k pace,” that is metabolically stable, but going just a little bit faster will tilt you into metabolic instability and the ensuing crash in your key physiological metrics.

That speed—the dividing line between metabolic stability and metabolic instability—is your SSmax speed. Your metabolic power output (as measured by, say, oxygen consumption) at that speed is your actual SSmax.

Differentiating these two concepts (speed at SSmax, and SSmax itself, the underlying biological construct) is useful because it allows us to separate your running economy from your SSmax. SSmax is fundamentally a metabolic power output; running economy determines how fast you can run at that metabolic power output.

How do you actually measure SSMax?

Steady-state max is the underlying biological phenomenon that underpins the metabolic dynamics we’ve just covered. But SSmax isn’t something you can directly measure in a lab. Instead, we need to choose a specific number to measure and choose a criteria to establish what, exactly, constitutes a “steady-state.”

A good analogy here is the idea of core temperature in medicine. Core temperature is a real biological phenomenon, in the sense that the heart, liver, etc., are all operating at a given ambient temperature, and this temperature is a good indicator of the status of your body overall.

However, getting a true core temperature measurement is extraordinarily difficult (though not impossible). In the clinic, doctors have a range of possible methods to estimate core temperature, like using a non-contact “temperature gun,” an oral thermometer, a rectal thermometer, or an ingestible capsule.

Each of these methods to estimate core temperature has different tradeoffs: temperature guns are fast but inaccurate, oral thermometers perform somewhat better but can be affected by the air temperature, rectal thermometers are accurate but invasive, and ingestible capsules are expensive.

Estimates of SSmax share many of the same properties: there are tradeoffs between accuracy and convenience.

Lactate threshold (LT2) is a quick but less-accurate estimate of SSmax

Most runners are familiar with threshold as a concept—in the popular conception, “lactate threshold” is “the speed at which blood lactate levels start rising dramatically.”

That’s an intentionally vague description, since I think most people have only this level of vague familiarity with lactate threshold.

There’s a few layers we need to peel back. First, lactate threshold testing uses a single-visit progressive treadmill test. A typical protocol might look like this:

1. Do 10 min of easy jogging on a treadmill, followed by a baseline lactate measurement
2. Complete a 3 min stage of treadmill running at an easy pace, then rest 20 seconds to take blood lactate measurement.
3. Increase the treadmill speed 0.5–1 km/hr, then complete another 3 min stage and measure blood lactate
4. Continue until it is too difficult to complete another 3 min stage.

The idea with this protocol is that blood lactate levels after each 3 min stage are a way of estimating what blood lactate levels would be if you ran continuously at that same speed.

There are many misconceptions about blood lactate testing, which are often the result of selective (or sometimes completely made up) plots of what blood lactate test results are “supposed” to look like. The reality is a little messier than a textbook might suggest.

Here’s what real lactate testing data look like:

These lactate curves all have more or less the same shape, but when overlaid on top of one another we can see how much variation there is from one athlete to the next:

These data happen to be from the excellent paper on the Nike Breaking2 athletes, some of the top marathoners in the world.

Two things are obvious from these plots. First, baseline blood lactate varies quite a bit from person to person, and second, blood lactate at the end of each stage increases as a function of speed in a curvilinear fashion.

Classical physiology sources make two additional claims (which to me are far less obvious, but we’ll save that for another day). These claims are:

1. There is a speed at which blood lactate levels become significantly greater than baseline levels (i.e. higher than the post-warmup baseline measurement).
2. There is a speed at which blood lactate levels are “sharply higher” and continue to rise rapidly at faster speeds.

There are a lot of methodological details I could cover here but I’ll save those for a separate article dedicated to single-session lactate testing, and we’ll just take those two claims as a given.

LT1 is not an estimate of SSmax

The first point—the speed at which blood lactate levels are “significantly greater” than the baseline, is the first lactate threshold, or LT1. “Significantly greater” than baseline is often claimed to be the first point at which blood lactate exceeds 2.0 mM, but a better rule is 0.5 mM above the baseline value.

You need to be very careful, especially when reading older physiology papers and textbooks, since LT1 is often just referred to as “the” lactate threshold.^[6] LT1 is not what most runners and coaches mean when they say “threshold.” LT1 is not an estimate of SSmax.

LT2 *is* an estimate of SSmax (but not a super accurate one)

The second point—the speed at which blood lactate levels are “sharply higher” than the previous speed, is the second lactate threshold, or LT2. This speed is what most runners mean when they say “threshold.”^[7] Moreover, LT2 is an estimate of SSmax.

Defining what constitutes a “sharply higher” level of lactate is controversial. This is where the often-quoted cut-off of 4.0 mM comes from (though it’s not the most reliable way to identify LT2, as our plots above clearly show).^[8]

In any case, the real weakness of LT2 is the fact that it is relying on a one-off lactate measurement at different speeds as a proxy for lactate levels at the same speed over time. Like a temperature gun, LT2 testing is fast and relatively cheap, requiring only a single visit to a lab, but it is not the gold standard for estimating SSmax.

Maximal lactate steady-state (MLSS) is a better estimate of SSmax

If you want to know whether lactate levels are stable over time, why not just run the same speed and take sequential blood lactate measurements? This is the idea behind the maximal lactate steady-state or MLSS.

Here’s what a MLSS test protocol looks like:

1. Do 10 min of easy jogging on a treadmill, followed by a baseline lactate measurement
2. Set the treadmill to a moderate pace and begin running.
3. Measure blood lactate levels after 5 minutes, then continue running at the same speed.
4. Continue to measure blood lactate levels at 10, 15, 20, 25, and 30 minutes.
5. Stop the treadmill after 30 minutes.
6. Two or three days later, return to the lab and repeat steps #1-5 at a speed 1 km/hr faster.
7. Repeat until you reach a speed where blood lactate levels increase by more than 1.0 mM from minutes 10 to 30.^[9]
8. The last speed before this final speed is your maximal lactate steady-state. 

Here’s a typical example of MLSS testing:

By the traditional criteria, the second speed (denoted *MLSS above) is the maximal lactate steady-state for this runner—the next speed is just above the 1.0 mM over 10–30 min rule.

The advantages and disadvantages of the MLSS approach are clear: there’s less reliance on “magic numbers” like 4.0 mM, and  you get a true measurement of whether you’re at a steady-state over time (at least according to blood lactate levels).

However, MLSS testing requires multiple lab visits: usually you need at least four or five 30 min runs, separated by an easy day, to get a good estimate of MLSS. ^[10]

Moreover, the more accurate of a measurement you want, the more lab visits you need: bumping up the treadmill by only 0.5 km/hr for each run would result in more finely-grained determination of MLSS, but could easily double the number of runs you need to complete. Lastly, because MLSS involves “back-tracking” to the last metabolically stable speed, it has a tendency to slightly underestimate your true SSmax.

Ingebrigtsen-style lactate measurements are very similar to MLSS testing

One of the reasons there’s been a resurgence in interest in threshold workouts has been the success of the Ingebrigtsen family, and especially Jakob Ingebrigtsen, who makes extensive use of lactate testing during workouts to calibrate his effort level.

Ingebrigtsen’s “Norwegian method” of double threshold sessions (based on techniques pioneered by Marius Bakken) is often explained using the language of LT2 testing, but its use of sequential lactate testing during long repeats (e.g. 4–5 x 2km, testing lactate levels after each rep) looks much more like MLSS testing than LT2 testing.

Indeed, the explicit goal in these workouts is to adjust workout pacing so that blood lactate levels are stable over time, ensuring that the entire workout is completed at a metabolic steady-state.

Critical speed (a.k.a. critical velocity or CV) is the best estimate of SSmax

The critical speed model (which already has its own article) is currently considered the gold-standard estimate of SSmax, and has the benefit of not requiring any special equipment, but it’s also the most arduous of the three common methods.

While MLSS requires multiple in-lab visits, critical speed testing requires multiple all-out races or time trials. Officially, for research purposes, you need a minimum of three all-out trials, but for training purposes you can usually get away with just two.^[11] This is the biggest shortcoming of critical speed: getting a gold standard estimate requires a huge disruption to training.

Lab-based studies usually use time to exhaustion on a treadmill at different speeds, though time trials and races over set distances will work as well. A typical treadmill protocol looks like this:

1. Do 10 min of easy jogging on a treadmill
2. Set treadmill to a speed you think you can sustain for about 5 minutes
3. Run at that speed until you are too fatigued to continue, then record the distance you covered
4. Two or three days later, return to the lab and repeat steps #1-3 at a speed you think you can sustain for about 8 minutes
5. Two or three days later, return to the lab and repeat steps #1-3 at a speed you think you can sustain for about 12 minutes

The idea with critical speed is as follows: if you make a plot of the speed you can sustain for a given distance (or a given time), using race distances that last 2–20 minutes or so, your performances fall on a hyperbolic curve, which asymptotically approaches a critical speed—the critical speed—which just so happens to be an excellent estimate of SSmax.

That process looks like this:

The superiority of critical speed testing over MLSS testing was demonstrated in a 2021 study by Rebekah Nixon and colleagues at the University of Exeter. I detailed this study in my critical speed write-up, but the main takeaways are in the image below:

Notice how running just a little bit slower than critical speed (“CS-”) generates a VO2 steady-state, even though at this speed, blood lactate gradually rises at a rate higher than the traditional 1.0 mM from 10–30 min rule that’s used for MLSS testing.

Critical speed is an excellent model for research, and it’s 100% the way I would estimate SSmax if I were doing a physiology study. But what about coaching? I actually do not find traditional critical speed testing useful, because it’s a huge disruption to training to try to fit in multiple all-out races or time trials every few weeks to monitor progress

You can predict critical speed reasonably well from 5k fitness

Among runners who are already fit enough to run a 5k between 14:00 and 25:00, there is a shortcut to a full-fledged critical speed test: you can just use a percentage of your current 5k pace to estimate critical speed, and in turn, SSmax pace.

This section is mostly a placeholder—I have a full write-up in the works on how to predict critical speed from 5k pace, but the short version is that ~96–98% of 5k pace is metabolically unsustainable (just above SSmax), and ~90–92% of 5k pace is metabolically sustainable (just below SSmax) for most runners. Note that these are approximations; I am still working on the full analysis!

It depends a bit on your fitness level, though, and the final write-up will include a calculator with confidence intervals. Stay tuned and sign up for my email list to find out when that comes out.

Don’t think of SSmax as “one hour race pace”

SSmax and its lactate-based estimates are often claimed to be “one hour race pace,” or, variously, 25min, 30min, 45min, or 75min race pace (or 10k/15k/HM race pace). While it’s true that studies on trained distance runners often find that surrogates for SSmax, like LT2 and MLSS, tend to average out to around the pace that the subjects in these studies can sustain for 45–75 minutes or so, this is not the right way to think about SSmax. Ditto for claims about critical speed or “CV” being “25 minute race pace.”

SSmax is, again, a boundary between metabolically stable and metabolically unstable intensities. The factors that determine SSmax are not the same as the factors that determine how long you can sustain SSmax, or sustain a true metabolic state (given the uncertainty over whether running exactly SSmax produces an actual steady-state; see below for details).

This is not just a minor difference in definitions, but a point of real distinction for using SSmax in training. It’s very important to identify when our goal is to improve SSmax, and when our goal is to improve how long you can sustain a metabolic steady-state.

Improving SSmax versus maintaining a steady-state for longer

As an example, let's say you are coaching a middle distance runner with indoor season bests of 1:37 (600m), 2:15 (800m), 3:09 (1000m), and 5:16 (mile). We can estimate her SSmax pace from her performances using a critical speed model, which gives an estimated SSmax pace of 5:54/mi. Could this athlete maintain 5:54/mi for 25 or 35 minutes? Probably not!

More importantly, this athlete doesn’t need to be able to sustain SSmax pace for 25 or 35 minutes. But, if she wants to run a good 1500m or 3000m steeplechase during the outdoor season, she definitely needs to improve her SSmax pace.

In contrast, let’s say you’re coaching a recent college grad with a 5k best of 14:30. After graduation, he trained for a marathon, following a plan from a popular training book, and ran 2:26, coming through halfway in 1:11. His problem is not SSmax—it’s his ability to sustain a high fraction of SSmax for a long time. In his marathon, he likely got beat by runners with a much slower SSmax pace than his own.

As a final example, let’s say you are coaching a runner who just missed last year’s Boston cut-off with a 3:27 marathon time. She’d like to run 3:20 to have a solid BQ, and her half marathon PR is 1:36. For her, the best bet is to improve both her SSmax pace and her ability to run at a high percentage of SSmax to have the best shot at her goal.

These three runners need very different training to achieve their goals, and differentiating between situations where we want to improve SSmax, endurance at SSmax (or a high percentage of it), or both, is very helpful for making long-term training plans.

Applying the concept of SSmax to training

One of the reasons I think SSmax is such a foundational concept in physiology is that it’s almost always the first lens I use to look at a given running intensity.

If someone wants to run 3:00 in the marathon, for example, it matters quite a lot whether they’re already fit enough to maintain a metabolic steady-state at 6:50/mi. If they can’t, no amount of long easy runs will help: we need to improve their steady-state capabilities first. In contrast, if someone has very good performance in the 5k and 10k, but struggles in the marathon, the problem is not SSmax—it’s something else (probably resilience!).

SSmax also explains why more aerobically fit runners have an advantage in tactical races on the track. If a tactical 10k is below SSmax for a top runner, they can run the final mile or so of the race nearly as fast as if they were completely fresh, since maintaining a metabolic steady-state does not substantially deplete your anaerobic energy reserves over time.

SSmax is also useful to think about with regards to individual training sessions. If the goal is to run at a steady-state—for example, in a Daniels-style cruise interval workout—you should feel like you’re at a steady-state throughout the workout. If the workout is getting progressively more difficult as you continue, you’re likely not at a metabolic steady-state, even if you’re hitting the “correct” pace. 

Conversely, if the goal is to be above SSmax (say, in a workout at 5k pace, or 8k to 10k pace for faster runners), you should be feeling a gradual ratcheting up of pressure, difficulty, and fatigue as the workout unfolds.

You don’t necessarily want to run exactly at your SSmax

SSmax is closely associated with critical speed, a.k.a. “critical velocity” or “CV,” which has grown something of a reputation as a magic speed that you should be hitting in workouts. Even though I think workouts at SSmax have a place in training,^[12] you generally don’t want to run exactly your SSmax pace if you’re trying to generate a predictable physiological response.

Why not? Well, it’s more useful to think of SSmax as a boundary between different physiological domains, as opposed to a magic intensity. SSmax is like a ridge between two valleys, not a line in the sand.

What do I mean by that? It doesn’t always make sense to run exactly at SSmax, for the same reason that it's difficult to balance a soccer ball on a narrow ridge: it’s going to roll down one way or the other, in an unpredictable fashion.

If you try to run exactly SSmax pace, you'll be just like this soccer ball: perched in an unstable position!

Exercise physiology research suggests the same is true for SSmax: if you run 3–4% below SSmax, you’ll be at a metabolic steady-state, while if you run 3–4% above SSmax, you’ll be in metabolically unsustainable territory. Running exactly at SSmax results in unstable results, though—different biological systems will slouch towards or away from metabolic stability in an unpredictable fashion.

Intuitively, this makes sense, because any lab-based measurement of SSmax is going to have some uncertainty and error associated with it—going back to our soccer ball analogy, tiny variations along the top of the ridge are going to affect whether the ball rolls down one way or the other.

If your goal is to create a consistent and predictable metabolic state inside your body, it often makes more sense to either run a bit faster than SSmax, or a bit slower than SSmax (see my article on critical speed for a discussion of the concepts of CS+ and CS-, which get at this exact idea).

Benefits of running just above and just below SSmax

From a training perspective, keep in mind also that there’s no guarantee that running at SSmax is the best way to improve SSmax. It’s not a bad place to start, considering the principle of specificity, but there are plenty of reasons to believe that one effective way to stimulate your body to improve SSmax is to intentionally exceed it to put your body into a bit of a “crisis” and force it to adapt.

Likewise, it’s also useful to run at speeds that are substantially backed off from SSmax, both so you can ensure that you are in a completely stable metabolic state, and because slowing down a little bit can let you rack up a lot more volume. For example, running 10mi at 80% of 5k pace is a totally manageable workout, while 10mi at 90% of 5k pace is nearly an all-out effort (even though both paces will be below SSmax).

I was recently on the Strength Running Podcast to make the case for doing workouts in the high-end aerobic domain, i.e. at speeds fractionally backed off from SSmax—check out that podcast here if you’re interested. The benefits of running a bit backed off from SSmax are part of the motivation for doing so-called “sub-threshold” training as well, which is popular among Norwegian-style double threshold advocates.

Recap

Steady-state max, or SSmax for short, is the boundary that separates metabolically sustainable and metabolically unsustainable speeds. Just below SSmax, your oxygen consumption, blood lactate, intramuscular acidity, and other biomarkers are stable over time; just above SSmax, all of these same biomarkers gradually start crashing towards their limiting values.

When most runners and coaches say “threshold,” they mean SSmax. The traditional (second) lactate threshold, or LT2, is a simple and quick estimate of SSmax, but it’s not the most accurate one. Exercise physiology research either uses maximal lactate steady-state (MLSS) or critical speed (CS, or CV, for “critical velocity”) as a gold-standard estimate for SSmax.

SSmax is not a “magic pace” that you should try to run at in all of your workouts: it’s a boundary that delineates what speeds generate a metabolic steady-state, and what speeds tilt you into metabolically unsustainable territory.

When you run right at SSmax, you’ll get unpredictable results—in part because different biological systems are slouching towards or away from equilibrium, and in part because any estimate of SSmax is bound to have some uncertainty associated with it.

If you want to generate a predictable response from your body, it’s better to run just a bit above or  just a bit below SSmax, giving yourself a buffer of about 3% or so to ensure you’ll get the kind of response you want from your body.

You also shouldn’t make the mistake of thinking that the only way to improve SSmax is to run exactly at SSmax. There are good reasons to do some workouts at speeds that are significantly backed off from SSmax, so you can rack up a lot of time running at a fast but metabolically sustainable intensity. Likewise, one way to stimulate an improvement in SSmax is to intentionally exceed it by a moderate amount, pushing your body into a bit of a crisis to generate an adaptation.

Finally, the factors that determine your SSmax pace are different from the factors that determine how long you can maintain a metabolic steady-state. Raising SSmax is a worthy goal in most cases, but isn’t always the top priority. Sometimes, you’ll want to focus on sustaining a metabolic steady-state for longer.

Learn more about the science of running

If you enjoyed this article, subscribe to my email list! It’s the best way to find out when I’ve got a new article on the science of running.

If you want to learn more about the science of running, I also have a book, Modern Training and Physiology for Middle and Long-Distance Runners, that shows you how to apply principles from physiology to train for events from 800m to the 10k. Admittedly, this book considers metabolic steady-states through the older lens of LT1 and LT2, but in my defense, most of the research on the modern view of on steady-state max hadn't been published yet when I wrote it!

Footnotes

---

^[1] And the term “VO2max” also shares the same inversion of terms: VO2max stands for “maximal oxygen uptake,” with maximal at the beginning, even though it’s pronounced with “max” at the end. Occasionally I have seen the term “maximal aerobic steady-state” which abbreviates to “MASS” or “MAS,” which is easily confused with the more common “mass” in the sense of body mass, and whose inverted form is, of course, right out.

^[2] There are some corner cases where changes in oxygen consumption do not reflect a change in the true metabolic cost of running, like when you are running out of glycogen and shifting to greater fat oxidation, but for a well-trained runner, SSmax occurs at a high enough %VO2max that the difference between the true energetic cost of running and oxygen consumption is more or less irrelevant—see this paper for an extended discussion.

^[3] It is a common misconception to think that the first lactate threshold—LT1, where blood lactate levels rise above their resting baseline for the first time—represents “the point at which lactate production outpaces lactate reuptake.” This idea is wrong because blood lactate levels are stable over time at speeds above LT1 but below SSmax—and in fact, sometimes they gradually decrease over time. If blood lactate is stable, even if it’s stable at a level above baseline, that means that lactate production and lactate reuptake must be in equilibrium. Lactate reuptake is a concentration gradient-driven process, meaning your rate of lactate reuptake depends to some extent on the concentration of lactate in the bloodstream. So, blood lactate concentrations need to rise over baseline for your body to be able to reach its maximum lactate reuptake capacity. Low concentrations of blood lactate don’t provide enough “pressure” to fully saturate your lactate transport capabilities

^[4] 5k pace is metabolically unsustainable for pretty much everyone with a 5k PR below 25:00. If that doesn’t apply to you, just imagine it’s 3k pace instead.

^[5] Why “usually”? There are some edge cases very close to SSmax where it seems to be possible to achieve a steady VO2 but a gradually rising blood lactate concentration; we’ll discuss this later in the section on MLSS.

^[6] The terminology around LT1 and LT2 is a huge mess. All of the following terms may refer to the phenomenon captured by LT1: LT, AnT, AeT, AT, GET, VT, VT1, T1. All of the following terms may refer to the phenomenon captured by LT2: LT (also), LTP, AnT (also), AT (also), LTP, OBLA, RCP, VT2. Confused yet?

^[7] For example, “T pace” from Daniels’ Running Formula is an estimate of LT2 pace.

^[8] Many physiologists literally just eyeball it. From one paper: “[LT1 and LT2 were] identified by visual inspection. Blood [lactate]-speed plots were reviewed blind by four of the coauthors and a consensus on the running speeds at LT and LTP was reached and recorded.”

^[9] Why 1.0 mM? Round numbers like this are always a good bet that someone just made it up. Indeed, that seems to be the case—a recent paper suggests a threshold of a 2.0 mM increase from 10 to 20 min performs better than a 1.0 mM increase from 10 to 30 min. By this latter criteria, the MLSS in the plot above should actually be the next speed up.

^[10] There is one lab that claims to have developed a one-session MLSS test that involves “titrating” the treadmill speed every few minutes; the papers describing this technique do not have very many citations outside of the research group that established it, so my hunch is that this single-visit MLSS protocol is not seen by other physiologists as a viable replacement for the standard multi-day procedure. But this is one of those reading-between-the-lines judgement calls that's difficult to make.

^[11] As with MLSS, there is a single-visit “three minute test” that claims to establish critical speed, though it is mostly used in cyclists. I’m very hesitant to endorse it for runners because it is an extraordinarily demanding test: you run all-out for three minutes, intentionally not pacing yourself at all—i.e. you run top speed flat-out from the start, and hang on as best you can.

^[12] I’m of the opinion that every speed has a place in training, so in this sense there aren’t any magic paces at all. In the case of SSmax, I often incorporate workouts at 10k to 8k pace with the runners I coach—starting on the slower end, then progressing down to the optimistic side of 8k pace. I’m fully aware that we are intentionally starting the workout in a steady-state, then tilting into metabolically unsustainable terrain by the end. Aside from the more mathematical benefits of providing endurance support to 5k pace and providing speed support to HM pace, this strategy of intentionally exceeding SSmax also puts the body into a bit of a crisis, hopefully creating a stimulus to increase the SSmax.

Related articles