Lactate Testing for Triathletes: Two Tests, Not One
Triathletes need separate bike and run lactate tests. Test-week sequencing, eight matching tolerances, a 9 bpm worked example, swim sampling pitfalls, strip budget.
Short answer: Lactate testing for triathletes means running two separate step tests — one on the bike, one on the run — because power, pace and heart rate do not line up across disciplines. Swimming is usually tracked with pace and heart rate instead of blood samples, since clean sampling in a pool is hard to control.
Lactate testing for triathletes refers to measuring blood lactate during an incremental step test to locate the intensities — LT1 and LT2 — that anchor training zones, done separately for each discipline because bike, run and swim load the body differently. A triathlete needs two lactate tests, not one: a bike step test and a run step test, held on separate days and analysed as two independent files. The reason is mechanical rather than philosophical. The bike test returns watts, the run test returns pace, and heart rate — the only number the two files share — does not sit in the same place in both. Swimming is the third discipline and the one where capillary sampling usually falls apart for logistical reasons, which is why most triathlon test plans quietly become two-discipline test plans.
Five terms matter for reading what follows: LT1 (the first sustained rise in lactate above resting baseline, roughly the aerobic threshold), LT2 (the point where lactate rises steeply, roughly the anaerobic threshold), the step test (the incremental protocol that produces the data both are read from), VLamax (a marker of maximal anaerobic glycolytic rate, which some tools estimate from the same test data), and the lactate curve (the fitted line connecting stage values that LT1 and LT2 are pulled from).
What follows is the sequencing, the variables you have to hold constant, a worked example of how far apart two files from one athlete can sit, and the steps that most often destroy the data before analysis begins.
Why bike numbers cannot set run zones
Take a fictional athlete who completes both tests inside one week with matched protocols. Both anchors below use fixed lactate values (2.0 and 4.0 mmol/L), chosen here only so the two files can be compared at the same point — not as a recommendation for which method to use.
| Item | Bike step test | Run step test | | --- | --- | --- | | Stage length | 5 min | 5 min | | Increment | 20 W | 0.5 km/h | | Stages completed | 7 | 7 | | Load at the 2.0 mmol/L anchor | 218 W | 4:52 min/km | | Heart rate at that anchor | 148 bpm | 157 bpm | | Load at the 4.0 mmol/L anchor | 268 W | 4:12 min/km | | Heart rate at that anchor | 168 bpm | 177 bpm |
*Fictional data set, constructed to show the size of the gap. It is not a typical value and your own gap may be larger, smaller, or point the other way.*
The interesting column is heart rate. In this file the two disciplines differ by 9 bpm at both anchors. If you build one five-zone set from the bike test and use it for running, every run boundary lands 9 bpm below where the run test put it — and in the five-zone model derived from this same data set, the zones around the upper anchor are 8 to 12 bpm wide. The offset is not a rounding error; it moves a session a full zone.
That is the whole argument for testing twice. Two files, two unit systems, two heart-rate sets, stored separately and trended separately.
The test week: Order, spacing, and what contaminates the second file
The second test of a round is the one that goes wrong, because it inherits fatigue from the first. Three working rules:
- Leave roughly 48 hours between the two tests as a working convention. Shorter gaps are sometimes unavoidable; if your week only allows something closer to 24 hours, the data may still be usable for within-file analysis, but record the actual spacing in the file — it is now part of your protocol, and shorter recovery windows make comparisons across rounds less reliable.
- Fix the order once and keep it forever. Bike-then-run and run-then-bike both produce internally consistent data. Alternating the order between test rounds does not, because the second test is always the compromised one and you have moved which discipline carries the compromise.
- Keep the 48 hours before the first test free of maximal work and brick sessions. A brick two days out is the single most common source of a first stage that starts too high.
Write down what you could not control: sleep hours, ambient temperature, altitude, time since the last meal. You are not going to correct for these mathematically. You are going to look at them when a test round comes back strange.
One practical note: a maximal step test and repeated capillary blood sampling carry real physical strain and a small puncture-related risk, so treat the decision to test as a personal health question, not only a training one. If you have a cardiovascular, metabolic or bleeding-related condition, take medication that affects heart rate or clotting, or are unsure whether a maximal test is appropriate for you, check with a doctor before testing. Treat the resulting numbers as measurement output to discuss with a coach or a qualified professional — not as a diagnosis or a prescription.
Match these eight variables, or your retest is a new baseline
Comparability is the entire value of testing more than once. These tolerances are a working convention we propose, not an official standard — their purpose is that you write them down once and repeat them.
| Variable | Tolerance for a comparable retest | | --- | --- | | Stage duration | Exact. No tolerance. | | Increment size | Exact. | | Starting load | Exact. | | Sampling second within the stage | ±10 s | | Time of day | ±60 min | | Ambient temperature | ±3 °C | | Carbohydrate intake timing before the test | ±30 min | | Meter and strip system | Same system; record the lot number |
Miss one of the first three and you have not repeated a test — you have created a new baseline and thrown away the trend line you were building. That is worth more than the honest inconvenience of rebooking a session that has to move to a different time of day.
Stage length: The choice that quietly breaks a season of data
Blood lactate lags the workload that produced it. Exercise physiology generally treats 3–5 minute stages as long enough for lactate to respond to a step change in load, without being so long that a slow drift unrelated to that stage's intensity creeps in. In practice, a 3-minute stage tends to sample a value that has not fully settled; a 5-minute stage tends to sit closer to a plateau. Both stage lengths are common in step-test protocols, both produce readable curves, and neither is wrong on its own terms.
What does not work is comparing across them. A 3-minute protocol lets the athlete reach a higher final load before the test ends, so the same athlete's curve appears to sit further right than it does under 5-minute stages. Put a March test with 3-minute stages next to an August test with 5-minute stages and the difference you see is mostly protocol, not adaptation. If you are unsure which to adopt, pick the one your analysis method assumes and stop thinking about it — the consistency matters more than the choice. The method question itself is a separate decision; the trade-offs are laid out in calculate lactate threshold: methods and formulas explained.
Swimming: Why the same protocol rarely survives contact with water
Many triathlon lactate protocols skip the swim, or cover it only briefly, for practical reasons. Several problems show up in the pool far more often than on a bike trainer or treadmill:
1. Dilution. Pool water on the fingertip or earlobe thins the sample. Drying on a wet swimsuit is how most swim data sets get wrecked — dry with a clean, dry towel, then wipe the first drop away before collecting. 2. The sampling gap becomes a variable. Stopping at the wall, drying, sampling and restarting typically takes something in the order of 30 to 60 seconds. Whatever gap you use, it is now part of your protocol and belongs in the tolerance table above. 3. Stages are distance-based, not load-based. A 5 × 200 m set gives you pace per 100 m as the load axis, and stage duration then varies with how fast you swim it — the opposite of what the bike and run protocols hold constant. 4. You need a second person. Sampling yourself on a pool deck with wet hands, a lancet and a meter is a two-hand job you do not have.
If you cannot solve those four, log swim sessions with pace per 100 m, stroke rate and heart rate from a water-rated device, and keep blood lactate for the disciplines where the sampling workflow is clean.
What an LT1/LT2 calculator needs from a two-sport athlete
An LT1/LT2 calculator takes a table, not an impression. Per discipline, per stage, you need:
- Load: Watts for the bike, pace or speed for the run
- Stage duration and the second at which the sample was drawn
- Heart rate at the end of the stage
- Blood lactate in mmol/L
- One resting value before the first stage, and one or two post-maximal values
Enter the bike file and the run file as two separate analyses. Merging them into one table produces a curve that describes no athlete — it interpolates between two unit systems.
The other constraint is coverage. A curve needs at least five stages, with two clearly below the bend and two clearly above it; if your final stage sits below the point where lactate rises steeply, the LT2 output is extrapolation rather than measurement. Four stages that all sit in the flat part of the curve will still return a number. That number is the fit's guess, not your threshold. The mechanics of turning a stage table into a fitted curve are covered in calculate lactate curve online: step-test input and threshold output.
"Is a lactate of 3.3 high?" and other questions a single number cannot answer
It is not a question that can be answered as asked, and the reason matters more than the non-answer. A lactate value without its workload is a temperature without a location.
Before 3.3 mmol/L means anything, it needs: the load it was measured at, the stage duration, how many seconds after the stage ended the drop was taken, which meter and strip lot produced it, where the athlete's own resting and previous-stage values sat, and what the preceding 48 hours looked like. Change any one of those and the same 3.3 belongs to a different curve.
The same applies to the race-distance version of the question — whether a 10K is run "at threshold." One athlete finishes a 10K in 32 minutes and another in 68, so the same distance sits at different points on their respective curves. The usable move is to compare your own race heart-rate and pace file against your own test file, and to take the interpretation of both to a coach or qualified professional rather than to a rule of thumb.
Watch estimates and measured values belong in separate columns
Check your device's documentation for how it derives any figure it labels a lactate threshold. If the number appears without a blood sample, it is a model estimate built from heart-rate and pace data, and it will move when the model's inputs move — including after a firmware update that changes the model.
Feature availability also shifts between models and firmware versions, so verify what your specific device does in its current owner's manual rather than in a forum thread from two seasons ago.
Keep estimated and measured values in two separate series, even in the periods where they happen to agree. A trend line that mixes them records changes in the estimation model as if they were changes in you.
What a two-discipline test season costs in strips and hours
No prices here — meter and strip prices vary by market and supplier, and you should check current figures with the seller. But the consumable count is pure arithmetic, and many triathletes underestimate it by roughly half.
| Item | Per test | Bike + run round | 3 rounds / season (example) | | --- | --- | --- | --- | | Resting sample | 1 | 2 | 6 | | Stage samples (7 stages) | 7 | 14 | 42 | | Post-maximal samples | 2 | 4 | 12 | | Subtotal | 10 | 20 | 60 | | Wastage at roughly 1 in 12 (clotted, short, contaminated) | — | — | ~5 | | Budget (example) | — | — | ~65 |
*Example calculation for a seven-stage protocol run three times per season; adjust the stage count and wastage assumption to your own protocol.*
A single-sport runner on the same three-round plan needs 30 strips in this example, before wastage. The triathlete needs roughly double, because the arithmetic comes from testing two disciplines instead of one — the ratio holds however you change the stage count.
Time works the same way, as a rough estimate: seven 5-minute stages plus warm-up and cool-down runs to about 55 minutes per test, so roughly two hours per round and about six hours of testing across a three-round season, excluding setup and travel. If you are weighing that against booking a lab or a mobile service, the input list to compare is in lactate test cost comparison: lab vs. home testing.
Frequently asked questions
Can I use my bike test to set run zones if I add a fixed offset? Any offset you apply is borrowed from someone else's physiology unless you measured your own gap — and measuring your own gap means running the run test anyway.
How often should a triathlete retest? The useful trigger is a change you actually want to detect, not a calendar interval. Testing at the boundaries of training blocks is a common convention. The binding constraint is comparability: a retest you cannot match to the previous protocol tells you less than no retest at all.
What does a lactate threshold test cost? The drivers are the meter, the strips (count them from the table above), the lancets, your time, and — for lab or mobile formats — the session fee and travel. Ask any provider for the current total in writing, including whether analysis and a written report are included.
Do I need a lab? Both lab and self-testing formats produce readable curves. What separates them is who controls the protocol, whether the ergometer or treadmill is calibrated, and whether anyone hands you an interpretation afterwards. Decide which of those three you actually need before comparing prices.
Before your next test week
- Pick the discipline order and the stage length now, and write both into a file you will still find in six months
- Book the two sessions at the same time of day, roughly 48 hours apart, with no brick session in the 48 hours before the first
- Count your strips against the table above and order the wastage margin too
- Note the strip lot number on the protocol sheet, next to the ambient temperature
- Keep the bike file and the run file separate from the moment of entry — including in whatever tool you use to fit the curve
Do that, and the two files stay what they are meant to be: two independent series you can trend across a season, rather than one curve that describes no athlete.