Understanding Critical Power for rowers: what it is and why it matters

Your training zones are only as good as the anchor you use to set them. Most rowers use heart rate, and heart rate is a useful signal, but it lags, it drifts with caffeine and sleep debt and heat, and it tells you nothing about the power you're producing. Critical Power is a better anchor. It's grounded in what your muscles can actually do, not in a proxy signal one step removed from the work itself.

What Critical Power is

Critical Power (CP) is the highest power output you can sustain indefinitely without progressive fatigue accumulation. In practice, below CP you can keep going for a very long time; fatigue reaches a steady state rather than accumulating progressively. Above CP, fatigue accumulates, and eventually you stop. There's a hard physiological boundary there, not a gradient.

The concept pairs CP with a second quantity called W' (pronounced "W prime"). W' is your finite anaerobic work capacity, the total amount of work you can do above CP before you're forced to drop below it. Think of it as a battery. Every second you spend above CP drains the battery. Every second you spend below CP recharges it. When the battery hits zero, you're done.

Together, CP and W' describe a power-duration curve. For any duration T, the maximum average power you can sustain is:

$P_{\max}(T) = CP + \frac{W'}{T}$

At very short durations, W'/T dominates and you can produce enormous power. At long durations, W'/T becomes negligible and your sustainable power converges on CP. This is why your 500m pace is so much faster than your 2k pace, and your 2k pace is faster than your 60-minute pace. The math is clean and predictive.

How CP differs from FTP and lactate threshold

FTP (Functional Threshold Power) is a practical field estimate: the highest average power you can hold for roughly one hour. Lactate threshold is a laboratory measurement of the exercise intensity at which blood lactate begins to accumulate faster than it can be cleared. Both are trying to describe the same general region of intensity: somewhere around the aerobic-anaerobic crossover point.

CP is more precise and more theoretically grounded than either. FTP assumes a 60-minute duration, which doesn't apply cleanly to shorter events. Lactate threshold requires a blood draw. CP sidesteps both: it's defined by the mathematical structure of the power-duration relationship itself.

In practice, CP tends to sit slightly above lactate threshold and slightly above FTP. A 2024 meta-analysis found a mean difference of around 12 watts between CP and maximal lactate steady state, though individual variation is wide — the gap can differ by ±30 watts or more depending on the athlete and test protocol. Even a systematic difference of a few watts matters when you're setting zone boundaries, because the error shifts every zone.

Estimating CP from timed pieces you've already done

You need two or more all-out efforts at different durations. The classic approach is three efforts: something around 3 minutes, something around 10 minutes, and something around 30 minutes. On an erg these map naturally to pieces you're already doing: a hard 1k, a 2k or 4k, and a 30-minute steady-state piece done flat out.

Record the average power (in watts) for each effort. Then you're fitting those data points to the CP model. The simplest version is to plot total work (power times duration) against duration, fit a line, and read off CP as the slope and W' as the intercept. In practice, you want at least three points to get a stable estimate, and the efforts need to be genuinely maximal; a hard 2k you backed off on will give you a CP that's too low.

Here's an example with rough numbers for a masters rower:

Duration    Avg Power    Total Work
  3 min       280 W        50,400 J
 10 min       230 W       138,000 J
 30 min       195 W       351,000 J

Fitted CP  ~ 183 W
Fitted W'  ~ 16,800 J

The fitted CP is lower than all three test powers because it's the asymptote the curve approaches at long durations. W' is the intercept, the finite reserve that elevates your power above CP for any finite effort.

Why CP is a better zone anchor than heart rate alone

Heart rate is an output. It reflects what your cardiovascular system is doing in response to the metabolic demand. But it responds slowly, it saturates in ways that vary person to person, and it's sensitive to factors that have nothing to do with training adaptation: dehydration, stress, a bad night of sleep.

Power is an input. It's the mechanical work you're doing right now. It doesn't drift. It doesn't lag. And because CP divides the power-duration curve into fundamentally different physiological regimes, training zones built around CP have a principled meaning: Zone 1 is comfortably below CP (fat oxidation dominant, fully aerobic), Zone 2 is approaching CP from below (high aerobic demand, manageable W' drain), and anything above CP is drawing down W' and must be managed accordingly.

The table below shows how zones map onto the CP model:

Zone	Intensity	Description
1	< 55% CP	Recovery / easy aerobic
2	55–75% CP	Long aerobic base
3	75–90% CP	Tempo / threshold development
4	90–100% CP	Near-threshold; sustainable for 20–60 min
5	100–110% CP	Just above CP; W' depleting slowly
6	> 120% CP	Hard anaerobic; W' depleting fast

Zones 1 through 4 are sustainable in the sense that W' is not meaningfully depleted. Zones 5 and 6 are above CP, where you're spending W', and you need enough recovery below CP to replenish it before the next hard interval. This gives you a principled framework for structuring interval workouts, not just a set of arbitrary percentages. The broader implication — that most training time should be spent well below CP to build the aerobic base that makes hard sessions effective — is why the case against training hard every day is a useful companion read.

Re-testing

CP and W' aren't fixed values. They change with fitness. A good aerobic base phase raises CP; sprint work raises W'. That means you need to re-test periodically, roughly every 6 to 8 weeks during a structured training block. You're probably doing the test pieces already. A 2k time trial, a 30-minute piece, a hard 4x1k set. If you record average power for each, you have the raw material to re-estimate CP and track how your physiology is responding to training.

Most rowers have more useful data in their erg logs than they're using. The split and the duration are there. Converting to watts and fitting the CP model is the step most people skip.

---

ErgToday does this automatically — try it free for 14 days at erg.today.