← All news and papers

Blood Lactate Testing: Science, Technology and AI

29 September 2026 · Article · 12 minute read · Paul Forrest

In shortBlood lactate testing shows where easy becomes hard for each athlete: the science of LT1 and LT2, testing well, and how AI turns results into guidance. Written for coaches, with every research claim linked to its source.

Blood lactate testing tells a coach where easy running or riding turns into hard work, for one athlete, in numbers. A few drops of blood from a fingertip or earlobe, taken at the end of each stage of a progressive test, trace a curve with two turning points: the aerobic threshold (LT1) and the anaerobic threshold (LT2). Set training zones from those two points and every session is anchored to the athlete’s own physiology rather than to an age chart or a guess.

This article explains why the test matters, what the science says about running it well, how technology has moved it from the laboratory to the track, and where artificial intelligence can turn a page of numbers into guidance that fits the athlete in front of you, without pretending to know more than the data does.

Why blood lactate testing matters

Endurance coaching runs on intensity. Too easy and the session does not stimulate change; too hard, too often, and fatigue accumulates until the athlete stalls or breaks down. Heart rate, pace and effort ratings all help, but each is an indirect reading of what the body is doing. Lactate is closer to the source.

Decades of research back this up. Lactate measures track endurance performance more closely than VO2max does (Jacobs, 1986). The speed at which lactate begins to accumulate predicted performance from 3.2 km to the marathon, and marathon runners raced only just above it (Farrell et al., 1979). For a coach, that means a lactate test answers practical questions directly: how fast can this athlete go before the effort stops being sustainable, how wide is their tempo range, and has the last training block moved the curve?

The science: what the lactate curve tells you

Lactate is not a waste product. It is a fuel and a signalling molecule, made and used throughout the body (Brooks, 2018). At easy intensities it is cleared about as fast as it is produced, so blood levels stay close to resting values. As intensity rises, production outpaces clearance and blood lactate climbs.

Plotted against pace or power, the readings form a curve that stays flat, bends, then climbs steeply. A major review identified 25 different ways of locating its turning points and placed them in a two-threshold framework (Faude, Kindermann and Meyer, 2009):

The methods do not agree exactly. In a study that tested 48 cyclists five times each, the modified D-max method was the most repeatable and the best predictor of performance, and the fixed 4 mmol/L threshold did almost as well (Heuberger et al., 2018). The same study found that an athlete’s thresholds varied by 3.4 to 8.1% from one test to the next. Two lessons follow: report thresholds as a range across methods rather than a single number, and treat small changes between tests with caution.

Running a test that you can trust

The protocol shapes the answer. Shorter stages put thresholds at faster paces, because lactate has not had time to settle (Bentley, Newell and Bishop, 2007). In trained cyclists tested with stages from 1 to 10 minutes long, the closest estimate of the maximal lactate steady state came from four-minute stages (Jamnick et al., 2018).

What skews the numbers

A lactate reading is only as good as the conditions around it. Sampling site, temperature, recent exercise and diet all move the curve (Jacobs, 1986). Four effects matter most in practice.

How technology has changed lactate testing

Lactate testing used to mean a laboratory, a bench analyser and a sports scientist. Handheld analysers changed that: a strip, a drop of blood and a reading on the spot, at the track or beside a turbo trainer (Pyne et al., 2000). The bottleneck moved from measuring to making sense of the measurements. Fitting a curve, applying several threshold methods, reading off heart rates and building zones by hand takes time, and it is easy to get wrong in a notebook.

Software closes that gap. A coach enters each stage’s lactate, heart rate and pace or power as the test runs; the curve builds on screen; the thresholds, the range between methods and the training zones are calculated the moment the last stage is saved. Good software also checks the data while it is being collected: a reading that falls while pace rises, two strips that disagree, a resting value that is too high, or a rep that was slower than the one before.

That is the approach taken in Coach Lactate Lab, the lactate testing tool in the Athleet.Ai Track and Field and Triathlon coach editions. It runs on a phone at the track and keeps working without a signal once it is open. It works out LT1 and LT2 with several published methods, shows the range between them, flags data that needs a second look and produces a printable coaching report. Readings are typed in from the meter today; connecting meters over Bluetooth is a later step.

Where AI adds value: personal, athlete-centric guidance

A threshold is a number. What the coach and athlete need is a decision: what changes in next week’s sessions, and why. This is where artificial intelligence earns its place, as long as it is built on the calculations rather than in place of them.

The guard rails matter as much as the features. Language models can produce confident text that is not supported by their input (Ji et al., 2023). In sport science that is not a small risk: a wrong threshold sends an athlete into the wrong zone for weeks. The design rule is simple: the numbers come from the calculations, never from the model. In Athleet.Ai’s testing tools the results are worked out with published methods, and low-confidence results are shown as ranges with the reason. Where AI writes the report, as in Coach Force Lab, every number in the text is checked against the calculations before anyone sees it; Coach Lactate Lab’s report is currently built from rules. The coach always has the final word, and the guidance is training advice, not medical advice.

A practical checklist for your next test

Frequently asked questions

What is a blood lactate test?

A blood lactate test measures lactate in a small blood sample, usually from a fingertip or earlobe, at the end of each stage of a test that gets progressively harder. The results form a curve that shows the athlete’s aerobic threshold (LT1) and anaerobic threshold (LT2), which are used to set personal training zones.

What is a normal lactate threshold?

There is no single normal value. Lactate is low at rest, and testers aim for 1.5 mmol/L or less before a test starts. The fixed 4 mmol/L value is a common marker of the anaerobic threshold, but an individual’s true steady state can sit well above or below it, which is why several methods are used and the athlete is compared with their own previous tests.

How often should an athlete do a lactate test?

Every 6 to 8 weeks, or at the start and end of a training block, is a practical rhythm. In trained cyclists, thresholds varied by 3.4 to 8.1% from one test to the next, so testing much more often mostly measures that variation.

Is a fingertip or earlobe sample better?

Either works if you keep to one. In one study of incremental exercise, fingertip samples read 0.66 mmol/L higher than earlobe samples on average, so the two should not be compared with each other.

Can AI interpret a lactate test?

AI can explain a lactate test in the context of the athlete’s event and history, and turn their zones into session targets. It should not calculate the thresholds itself. The safest design calculates thresholds with published methods first, then lets AI write guidance that is checked against those numbers, with the coach making the final decision.

Sources