Critical Swim Speed: Setting Swim Intensity Without a Coach on Deck

Most age-group triathletes train in the pool by feel. That works up to a point, but feel drifts. A "hard" set on a tired Thursday isn't the same as a hard set on a fresh Monday, and it's very easy to spend months swimming at one comfortable, in-between effort. Runners and cyclists have long used test-derived zones to fix this. Swimmers have critical swim speed, which needs nothing more than a pool, a clock and two hard efforts.

This article explains what CSS actually measures, how trustworthy it is, how to run the test, and how to use the result without overreaching. It's the measurement side of the companion article "Six Months, Same Pace," which looks at why age-groupers plateau.

What critical speed is

The idea comes from a general model of endurance performance. If you plot distance swum against the time it took for several maximal efforts of different lengths, the points fall close to a straight line. The slope of that line is the critical speed: in theory, the speed at which the relationship would continue indefinitely. The intercept is sometimes read as a finite reserve of work you can do above that speed, often written as D′ (D-prime), the swimming equivalent of W′ in cycling.

TriForward's articles on W′ and critical power used the same model for cycling. CSS is its swimming version, and much of the same logic applies. Efforts above critical speed drain a limited reserve. Efforts below it can be sustained for much longer.

In practice, you don't need five time trials. Martin and Whyte (2000) tested eight elite triathletes (five men, three women) over 100, 200, 400, 800 and 1,500 m. Critical swimming velocity came out "similar regardless of the combination or number of time trials used." Their conclusion was that it can be calculated from any two time trials. That is why the common field version uses just two distances, usually 400 m and 200 m.

CSS is not an all-day pace

The same Martin and Whyte study is also the most important caution about CSS. The researchers compared critical swimming velocity with the velocity at lactate threshold, measured with a graded set of paced 300 m swims. In every subject, CSS was faster: 1.23 m/s against 1.15 m/s on average. Blood lactate at CSS was 3.0 mmol/L against 1.9 mmol/L at threshold. The authors concluded that because CSS is faster than lactate threshold, it is unlikely to be sustainable indefinitely, and that the "notion" of critical velocity as an indefinite pace should be re-evaluated.

Eight people is a small sample, and these were elite athletes. But the finding fits the wider critical-power literature, which TriForward covered for cycling in "Not All Kilojoules Are Equal." Critical speed sits near the boundary between the heavy and severe intensity domains. It's hard, it's sustainable for a limited time, and it isn't an "easy aerobic" marker.

The practical upshot is to use CSS as the top of your threshold work, not the pace for long continuous swims.

How reliable is the test?

A testing protocol is only useful if it gives similar answers when nothing has changed. The best recent data comes from Scott, Burden and Dekerle (2023), who tested 32 national-standard swimmers across all four strokes, 10 of them in front crawl, using 200 m and 400 m time trials on repeat occasions.

• The time trials themselves were very repeatable: coefficient of variation under 2%, with high reliability coefficients.
• Critical speed calculated from them was reliable: coefficient of variation under 3.4%.
• D′, the reserve estimate, was not reliable: coefficients of variation between 13% and 45%.

Two practical lessons follow. First, the 400/200 approach is a sound way to estimate CSS. Second, ignore the D′ number that some apps and calculators report alongside your CSS. It's too noisy to track.

There's also a lesson in the 3.4% figure, and this is TriForward's inference rather than something the study reported. If repeat tests in stable swimmers vary by around 3%, a change in your CSS of 1-2% between tests may just be noise. Look for consistent shifts across two or more tests before concluding something has worked or failed. These were national-standard swimmers. Less experienced swimmers, who may pace time trials less evenly, could plausibly show more test-to-test variation.

Running the test

A practical protocol for most age-groupers:

1. Warm up thoroughly. 400-600 m easy, including some drills and a few short efforts building to fast.
2. Swim 400 m flat out, aiming for even pacing. Going out too hard is the most common mistake and makes the result less repeatable.
3. Recover fully. Swim easy for at least 10 minutes, or do the 200 m on a separate day. Rushing this makes the 200 m slow and the calculated CSS artificially high. That last point is arithmetic, not a published finding.
4. Swim 200 m flat out, again aiming for even pacing.
5. Record both times, and ideally your stroke count on a middle length of each.

Use the same pool, pool length and conditions each time you retest. Turns make 25 m pools faster than 50 m pools, so don't compare results across them.

Doing the calculation

CSS in metres per second = (400 − 200) ÷ (400 m time − 200 m time), with times in seconds.

Worked example (TriForward's own arithmetic, not from a study):

• 400 m in 7:20 = 440 seconds
• 200 m in 3:20 = 200 seconds
• Difference = 240 seconds
• CSS = 200 ÷ 240 = 0.833 m/s
• Pace per 100 m = 100 ÷ 0.833 = 120 seconds = 2:00 per 100 m

Notice what this implies. The athlete swam the 400 m at an average of 1:50 per 100 m, but their CSS is 2:00 per 100 m. That's consistent with critical speed being a pace sustainable well beyond 400 m but for a limited time, not the pace of a maximal 400.

Turning CSS into training

The most direct evidence for CSS-based training comes from Piatrikova and colleagues (2020). They worked with 12 elite and national-level swimmers over 15 weeks on a reduced-volume programme of no more than 30 km per week, roughly 25% less than usual. High-intensity interval sessions were prescribed from each swimmer's own critical speed and critical stroke rate, a stroke-rate equivalent derived the same way. Critical speed improved by about 5.4%, stroke rate at critical speed by about 6.4%, and main-event performance by roughly 1-2%, despite the lower volume. D′ fell by about 25%, which fits with the reliability concerns above and with the programme's aerobic emphasis.

That's a small, uncontrolled study in a very different population from age-group triathletes. But it shows the principle working: individualised paces from a simple test, used to structure intervals, can drive improvement without adding volume.

How to translate CSS into sessions is largely coaching convention rather than tested prescription. A common framework:

• CSS sets: repeats of 100-400 m at CSS pace with short rest (roughly 10-20 seconds). This is your threshold work. Total CSS-pace volume might start at 800-1,000 m in a session and build gradually.
• Above CSS: short repeats (25-100 m) faster than CSS with longer rest, for speed and to practise holding form at higher stroke rates.
• Below CSS: steady aerobic swimming noticeably slower than CSS, for base, technique consolidation and recovery. Much of your total distance should sit here.

These bands aren't validated thresholds. Treat them as a starting structure, and adjust if CSS sets feel unsustainable after a few repeats. If they do, your test may have overestimated CSS, perhaps because you didn't recover fully before the 200.

Critical stroke rate: an optional extra

The Piatrikova study also used critical stroke rate, calculated by plotting the number of strokes against time across trials, much as speed is plotted against distance. For age-groupers, the simpler version is to note your stroke rate or stroke count at CSS pace and use it as a technique check. If you can hit CSS pace only by spinning your arms faster than at your last test, your fitness may have improved while efficiency has not. TriForward's "Stroke Rate vs. Distance-Per-Stroke" covers that trade-off in detail.

What the gap between your two times tells you

The two raw times carry information beyond the CSS number itself. TriForward offers this as a practical interpretation, not a validated diagnostic.

• If your 200 m pace is much faster than your 400 m pace, you have speed but fade over longer distances. Your CSS will be relatively low compared with your 200 m. More CSS-pace and aerobic work is likely to pay off.
• If your 200 m and 400 m paces are very close, you can sustain what you have but have little extra speed. Your limiter may be top-end speed or, more often in triathletes, efficiency: you can't go faster because your stroke can't use more power. Short fast repeats with good form, and technique feedback, are the priority.
• If your 400 m felt uncomfortably uneven, fast early and fading badly, treat the result as provisional and retest after practising even pacing.

Use this as a prompt for what to look at next, not as a firm conclusion. A single pair of trials can't separate fitness from pacing skill.

Using a benchmark set between tests

Full tests are tiring, and you don't need to run one every week. Many coaches use a repeatable benchmark set between tests, for example 8 × 100 m at CSS pace with 15 seconds' rest, swum once every two weeks. If the set gets easier at the same pace, with lower perceived effort, a lower stroke count or less fade over the last few repeats, your CSS has probably moved and it's worth retesting. If it gets harder, you may be tired or under-recovered rather than less fit.

This hasn't been studied as a monitoring tool in triathletes, and it isn't a substitute for a proper test. But it keeps your intensities honest between tests, using a session you'd be doing anyway.

How CSS relates to race pace

CSS is a training and tracking tool, not a race target. A few points, mostly inference:

• Sprint distance: a 750 m race swim may be close to CSS for a well-trained swimmer. But Peeling and colleagues (2005) found that swimming the first leg at 80-85% or 90-95% of maximal speed, rather than all-out, led to faster bike splits in nine male triathletes. Racing at your full CSS ability isn't necessarily optimal.
• Olympic and longer: you'll swim progressively further below CSS. A full-distance swim is far longer than critical speed can be sustained.
• Open water and wetsuits change the numbers. A 2021 review found wetsuits improved performance by up to 11%, varying with ability and suit, while sighting, chop and congestion slow you down. Your pool CSS won't translate cleanly to an open-water pace.

Use CSS in the pool to set intensities and to see whether you're getting fitter. For races, combine what CSS tells you about your fitness with experience of what you can hold in open water and still bike well.

Common mistakes

• Going out too fast in the 400 m, then fading. This lowers reliability and can skew CSS.
• Not recovering fully before the 200 m. A tired 200 m narrows the gap between the two times and inflates CSS.
• Treating CSS as a comfortable long-swim pace. It's above threshold for most people, as the elite triathlete data showed.
• Chasing small changes. With a typical variation around 3%, a 1-2% change is within noise.
• Tracking D′. It isn't reliable enough to be useful.
• Testing in different pools or pool lengths and comparing the results.
• Testing when tired. Put the test after an easier day, not straight after a hard training block, unless you're deliberately measuring fatigue.

A note on who these studies tested

The reliability data come from national-standard swimmers whose sex split wasn't reported in the abstract, and the training study from elite and national-level swimmers. Martin and Whyte's comparison included three women among eight elite triathletes. None of this was done in adult age-group triathletes learning to swim efficiently. That population is likely to pace time trials less evenly and to respond to training differently. The underlying calculation is robust. The precise reliability figures and training effects may not transfer exactly. Testing regularly in consistent conditions is the best way to find out how the method behaves for you.

References

• Martin L, Whyte GP. Comparison of critical swimming velocity and velocity at lactate threshold in elite triathletes. International Journal of Sports Medicine, 2000;21(5):366–368. doi:10.1055/s-2000-3786
• Scott BE, Burden R, Dekerle J. Stroke-specific swimming critical speed testing: balancing feasibility and scientific rigour. Journal of Human Kinetics, 2023;90:239–251. doi:10.5114/jhk/170882
• Piatrikova E, Gonzalez J, Willsmer N, Sousa A, Williams S. Individualizing training in swimming: evidence for utilizing the critical speed and critical stroke rate concepts. International Journal of Sports Physiology and Performance, 2020;15(5). doi:10.1123/ijspp.2019-0546
• Peeling P, Bishop D, Landers GJ. Effect of swimming intensity on subsequent cycling and overall triathlon performance. British Journal of Sports Medicine, 2005;39(12):960–964. doi:10.1136/bjsm.2005.020370
• Quagliarotti C, Cortesi M, Gatta G, Bonifazi M, Zamparo P, Baldassarre R, Vleck V, Piacentini MF. Wetsuit use during open water swimming. Does it "suit" everybody? A narrative review. International Journal of Sports Physiology and Performance, 2021;16(9):1217–1224. doi:10.1123/ijspp.2020-0808