Critical swim speed: the only pool test worth repeating
In the pool, the two metrics you trust on land both stop working.
Everything you use to judge intensity on land degrades in water.
Heart rate is suppressed — the horizontal position, the hydrostatic pressure on the chest and the cool water combine to give a heart rate several beats lower than the same metabolic effort on a bike. Wrist optical sensors, meanwhile, are being asked to read blood flow through a wrist that is moving through water at speed, which is close to the worst case an optical sensor can face. And perceived effort in swimming is dominated by technique: on a day when your catch feels good, hard work feels easy.
What is left is the clock. Which turns out to be enough.
The test
Swim 400 m as fast as you can hold. Rest properly — ten minutes, easy. Then swim 200 m as fast as you can hold.
Critical swim speed is the difference:
CSS (m/s) = (400 − 200) / (t400 − t200)Both distances in metres, both times in seconds. If your 400 takes 6:00 (360 s) and your 200 takes 2:50 (170 s), CSS is 200 / 190 = 1.05 m/s, which is 1:35 per 100 m.
That is the pace you can theoretically sustain without accumulating fatigue — the swimming equivalent of a functional threshold. The method comes from Wakayoshi and colleagues in 1992, and it has held up for thirty years because the arithmetic removes exactly what you want removed: the fast-start component that both trials share.
Why the subtraction is the clever part
Your 200 time and your 400 time both include your anaerobic contribution — the sprint you can spend once. Subtracting one from the other cancels most of it, leaving the sustainable component. That is why CSS is more stable than either time trial alone, and much more stable than “the pace I did my last 1,500 at”.
It also means the test is only as good as your pacing. If you go out too hard in the 200, or sandbag the 400, the subtraction inherits the error. Swim both honestly, on the same day, in the same pool, and the number is repeatable within a second or two per 100 m.
What to do with it
Zones fall out of it directly. Easy aerobic work sits 8 to 12 seconds per 100 m slower than CSS; threshold intervals sit at CSS or a second or two under; VO₂max work sits 4 to 6 seconds per 100 m faster. The great virtue of this is that your interval targets stop being guesses and start being derived from a measurement you took yourself.
Retest every six to eight weeks. Typical values run around 2:00–2:20 per 100 m for a recreational swimmer, 1:30–1:50 for a trained age-grouper, and 1:10–1:30 for competitive swimmers — but as with everything here, the useful comparison is with your own last test, not with the table.
Two or three seconds of CSS improvement is a meaningful block of training. It is also, conveniently, about the smallest change the test can resolve.
Swim Analytics does the subtraction and builds the zones from it.