CdA Calculator — Drag Area from Power and Speed

What CdA Is

CdA is drag coefficient multiplied by frontal area, in square metres. It is the single number that describes how hard you are to push through air, and above about 25 km/h it is the dominant thing standing between your power and your speed.

At 45 km/h on the flat, around 85–90% of the power reaching the wheel is going into moving air out of the way. Rolling resistance is most of the rest. This is why aerodynamics gets the attention it does in time trialling, and why a 0.01 m² improvement is worth more than a kilogram off the bike on anything that is not a climb.

The Model

This is not a fitted curve. It is a force balance, which means the only uncertainty in the answer is the uncertainty in what you put in:

P_wheel = ½ρ·CdA·v_air²·v_ground + Crr·m·g·cos θ·v_ground + m·g·sin θ·v_ground P_meter = P_wheel / (1 − drivetrain loss)

Rearranged for CdA:

CdA = 2 · (P_wheel/v_ground − Crr·m·g·cos θ − m·g·sin θ) / (ρ·v_air²)

Note which velocity appears where. Drag depends on air speed** squared — your ground speed plus any headwind — while the power it costs is that force times **ground speed. Getting those two confused is the most common error in home-made spreadsheets, and it makes every windy run wrong.

Air Density Is Not 1.225

Most CdA calculators hardcode sea-level standard density, 1.225 kg/m³. That is a choice that quietly destroys the measurement, because a 5% error in ρ is a 5% error in CdA — and 5% of a 0.25 m² CdA is 0.0125 m², which is larger than almost any position change anyone is trying to detect.

So this page computes it:

ρ = p_dry/(R_dry·T) + p_vapour/(R_vapour·T)

from your pressure, temperature and humidity. The implementation returns exactly 1.2250 kg/m³ at ISA sea level, 15 °C, dry — which is the check that it is wired up correctly.

Three things that follow, and two of them surprise people:

Altitude is huge. At 2,000 m and 15 °C, density is about 0.957 kg/m³ — 22% below sea level. Measure a position at altitude, assume 1.225, and you will report a CdA 22% too high.

Heat matters more than you think. 35 °C at sea level gives about 1.13 kg/m³, roughly 8% below a cool morning. The same rider at the same CdA will be measurably faster in the heat.

Humid air is *less* dense, not more. Water vapour has a lower molar mass than the nitrogen and oxygen it displaces, so adding moisture at constant pressure makes air lighter. "Heavy humid air" is a real sensation and a wrong physical description. The effect is small — about 0.6% from dry to saturated at 15 °C — but it is in the right direction only if you compute it rather than guess.

One input trap: enter station pressure, the actual pressure where you are, not the sea-level-corrected figure your weather app shows. Those are the same number only at sea level. Entering 1013 hPa while standing at 1,500 m overstates density by roughly 18%, and that error lands directly on your result.

Crr Is the Weak Link

Rolling resistance is the one term you cannot measure from the same test, so it enters as an assumption:

At 45 km/h, moving Crr from 0.004 to 0.006 shifts about 12 W out of the drag term, which is roughly 0.01 m² of apparent CdA. So a CdA quoted without a stated Crr is not a measurement of anything.

The practical consequence: do not compare your CdA to a published figure unless you know what Crr that figure assumed. Do compare your own runs to each other on the same surface and tyres, where the Crr error is a constant offset and cancels out of the comparison.

The Protocol Decides Whether the Number Means Anything

A single run gives you one number with that moment's wind baked into it. The arithmetic here is exact; the measurement is not, and the gap is entirely in how you ride.

What works:

1. Flat, straight, closed or quiet road. Any gradient error becomes a CdA error, and a 0.5% gradient you did not notice is about 20 W at 80 kg and 45 km/h. 2. Out and back, several times. A steady wind inflates one direction and deflates the other by nearly equal amounts in the power domain. Averaging the pairs cancels most of it. One downwind run will flatter you by a lot. 3. Hold speed steady, not power. Acceleration is a term this steady-state model does not have. Smooth speed over a fixed segment, not a smooth power file. 4. Same bottles, same clothing, same helmet position. You are measuring a system, and a flapping jersey is part of it. 5. Change one thing at a time. Two changes and one number tells you nothing about either.

What does not work: using a ride from Strava where the speed was variable, the route had turns, and the wind was whatever it was. That produces a number, and the number is noise.

Reading the Answer

These scale with body size — a 1.90 m rider will not reach a 1.65 m rider's number, and should not try to. If your result is outside every bracket, the inputs are wrong before the position is.

What a Change Is Worth

A CdA figure on its own does not tell you whether a change is worth making. The useful translation is time and watts, which is what the last block on this page does: it solves the speed your power buys at each CdA and differences the finishing times.

For a typical rider at 300 W over 40 km, dropping CdA by 0.01 m² is worth somewhere around 45–50 seconds, or about 12 W at the same speed. That is the frame to judge a skinsuit, a helmet, or a position change in — not the CdA delta itself, which means nothing to anyone outside the discussion.

For the power side of the same question, the watts per kilogram calculator and the FTP calculator cover what you are producing; this one covers what happens to it.