VLamax: the number that explains why two riders with the same FTP race differently

Two riders turn up to the same 40-minute hill climb with the same functional threshold power — 280 W each, tested the same week. One of them wins a bunch sprint most weekends. The other gets dropped in every criterium and then rides away from everybody on a long climb. Their FTP says they are the same rider. Nothing else about them agrees.

VLamax is the number that coaching circles reach for to explain that gap. It stands for the maximal rate at which your muscles can produce lactate through glycolysis — in other words, how fast your fast, oxygen-independent energy system can run when you open it fully. Where VO2max describes the ceiling on your aerobic engine, VLamax is supposed to describe the size of the glycolytic one.

It is a genuinely useful idea. It is also a considerably shakier measurement than the enthusiasm around it suggests, and you should know both halves before you build a season on it.

What the number means

VLamax is reported in millimoles of lactate per litre per second. Values commonly quoted for trained cyclists run from roughly 0.3 to 0.9 mmol·L⁻¹·s⁻¹, with endurance specialists towards the bottom and track sprinters above the top of that range. A higher number means glycolysis can spin up faster and harder.

The reason this matters for endurance is not obvious until you see the model behind it. In the metabolic framework that VLamax comes from — described by Mader and Heck in the 1980s, and the same lineage that gave us the 4 mmol/L threshold convention — your threshold is not an independent property. It is the intensity at which lactate production and lactate clearance balance. Raise the maximum rate of production and, holding oxidative capacity constant, that balance point moves down.

That is the claim in one sentence: for a given aerobic engine, a bigger glycolytic engine gives you a lower threshold as a fraction of VO2max, a faster sprint, and a higher carbohydrate burn rate at any given pace. It explains why a strong sprinter can have an unimpressive fractional utilisation, and why a pure time-triallist so often has no finishing kick at all.

Two riders, same FTP

This is why the metric caught on with coaches. If you only know FTP, these two riders look interchangeable and will be handed the same plan. If you know the shape of their power curve as well, they obviously are not. VLamax is an attempt to put one number on that difference and connect it to a mechanism rather than just a description.

How it is measured

The test itself is short and unpleasant.

The arithmetic looks trivial: change in blood lactate, divided by time. The trouble hides in that last step. The first seconds of an all-out sprint are fuelled mostly by stored phosphocreatine rather than glycolysis, so the convention is to subtract a fixed "alactic time span" — usually 3.5 seconds — before dividing. On a 15-second sprint, that constant is nearly a quarter of the denominator. Work published in 2023 showed the alactic period is not a constant at all but varies between individuals, which means the assumption quietly changes the answer by a meaningful amount depending on who you test.

There is a second wrinkle. Measured in the same athletes, VLamax comes out different in cycling than in running. It is a property of the exercising muscle mass and the mode, not a whole-body constant, so a number from one activity does not transfer to another.

Which direction do you want it to go?

This is the part that makes VLamax useful as a planning idea even if you never measure it. The answer depends entirely on your event.

A flat 40 km time trial or a long climb wants a low number. Every millimole of lactate produced at race pace is carbohydrate spent and clearance capacity occupied. A lower VLamax pushes your maximal metabolic steady state up as a fraction of VO2max and reduces the carbohydrate cost of any given pace, which matters more the longer the event runs.

A criterium, a track event or any race decided in the last 200 metres wants a high number. You cannot sprint without glycolysis, and grinding your glycolytic capacity down in the name of a tidier threshold is an excellent way to arrive at the finish with nothing left to throw.

A road race or a gran fondo wants both, which is the same awkward compromise every road cyclist has always faced. The framework at least makes the trade-off explicit rather than pretending you can maximise everything at once.

Notice that none of these levers is new. They are the sessions you already know, sorted by which energy system they push on. That is arguably the honest value of VLamax as a concept: it gives a coherent reason for why a time-triallist should stop sprinting in February, and why a track rider should not do a twenty-hour base block. Moveee's power profile will not measure your glycolytic rate — nothing without a lancet can — but the gap between your 15-second power and your 40-minute power is the same information in cruder form, and it is free to look at.

Where the evidence actually stands

Now the part that usually gets left out.

The model behind it has been tested. The metric itself, much less.

The metabolic simulation that VLamax comes from is not fringe work — it is the same modelling tradition that produced the 4 mmol/L threshold and has been used to calculate maximal lactate steady state from a lactate test rather than from a series of 30-minute rides. Comparisons of calculated against experimentally measured steady state have found reasonable agreement, and the calculation has been shown to be repeatable in amateur cyclists. That is a real body of validation for the framework.

What is much thinner is evidence about VLamax as a standalone training target. Reliability of the 15-second test has been examined in small samples with results that are best described as acceptable rather than excellent, and a 2024 study looking at the repeatability of sprint power, blood lactate accumulation and phosphagen contribution together found the derived metabolic measures noisier than the raw power output feeding them. Small samples, few laboratories, short follow-ups.

There is no equivalent of the VO2max literature

Compare that with VO2max, which has a century of measurement behind it, or with the critical power model, which has been tested against direct physiological criteria in dozens of independent labs. Both of those have known error bars, known failure modes and large prospective datasets. VLamax has a plausible mechanism, a handful of reliability papers, and a great deal of coaching enthusiasm. Our pieces on VO2 max and Critical Power versus FTP are the right comparison points if you want to see what a well-validated metric looks like.

What that means for you

• Do not re-plan a season around a single VLamax test. One number from one sprint on one day, with an assumed constant baked into the denominator, is not a foundation.
• Trends in your own repeated tests are more defensible than absolute values. The same rider, same lab, same protocol, same time of year is a comparison that survives most of the measurement problems.
• Use the concept without the blood. "Am I training the system my event actually needs, or the one I enjoy?" is the useful question, and you can answer it from your power curve and your race calendar.
• Be suspicious of anything sold as a VLamax estimate from power data alone. Lactate kinetics are not recoverable from watts. Those numbers are a model on top of a model.

The honest summary

VLamax describes something real: riders differ enormously in how fast they can run glycolysis, and that difference shapes threshold, fuel use and sprinting in ways FTP alone cannot capture. As a way of thinking about why two riders with the same threshold race like different species, it earns its place alongside LT1 and LT2 in your mental model.

As a number to test, track and optimise, it is not yet on the same footing as the metrics it is often listed beside, and anyone telling you otherwise is ahead of the evidence. Treat it as a good idea with a measurement problem rather than a solved instrument, and you will get the useful part without the overconfidence.