What the quarter-mile correlations actually predict
A quarter-mile prediction is a statement about power-to-weight, and nothing else. Feed in race weight and flywheel horsepower and you get back two numbers: elapsed time, measured from the moment the front wheel rolls out of the stage beam to the moment it breaks the beam 1,320 ft later, and trap speed, the average speed over the last 66 ft of that run.
Those two numbers behave very differently, and that difference is the most useful thing on this page. Trap speed is almost purely a function of power-to-weight, because by the time the car reaches the traps the launch is ancient history and the engine has spent the whole run doing work against inertia and drag. Elapsed time carries everything else: how well you left the line, whether the tyres hooked, how many times you shifted, how much of the run you spent sideways. Two cars with identical trap speeds and half a second between their ETs are telling you exactly where the slower one is losing.
That is why racers treat trap speed as the honest broker. If your ET is a long way off this prediction but your trap speed matches, the engine is fine and the problem is at the start line. If the trap speed is down too, you have less power than you think — and the trap speed horsepower calculator will tell you how much less.
Why the formula is a cube root
The cube root is not arbitrary. Model the car as a machine that delivers roughly constant power P to the road for most of the run. Power is force times velocity, so as the car speeds up the available thrust falls. Integrate that and you get distance rising with time to the power of three halves — which inverts to time rising with distance to the power of two thirds, and to elapsed time scaling with the cube root of mass divided by power.
The same model gives trap speed scaling with the cube root of power divided by mass. The two formulas are the same physics written twice, which is why ET × MPH is very nearly constant for a given constant pair: with the Fox numbers, 5.825 × 234 = 1,363, so a car running 117 mph should be around 1,363 ÷ 117 = 11.65 seconds. That identity is a fast mental check on any timeslip.
What the constants absorb is everything the constant-power model leaves out: the first 60 ft where the car is traction-limited rather than power-limited, aerodynamic drag rising with the square of speed, driveline losses, and shift times. Those effects are large, but they scale in a broadly similar way across ordinary cars, so a single fitted constant works surprisingly well over a wide range.
Two constants are in common use. Roger Huntington published the 6.290 and 224 pair in the 1960s, fitted to the cars of that era on bias-ply tyres. Geoffrey Fox later published 5.825 and 234, which fit modern chassis, modern rubber and modern gearing considerably better. On the same 3,200 lb, 400 hp car the two disagree by 0.93 seconds and 5 mph — a reminder that neither is a law of nature.
Worked example: 3,500 lb and 400 hp
Take a typical street-strip car: 3,500 lb across the scales with the driver in it, and 400 flywheel horsepower.
- Divide weight by power. 3,500 ÷ 400 = 8.75 lb/hp.
- Take the cube root. 8.751/3 = 2.0606. (Check it: 2.0606 × 2.0606 × 2.0606 = 8.750.)
- Multiply by 5.825 for the ET. 5.825 × 2.0606 = 12.00 seconds.
- Divide 234 by the same root for the trap speed. 234 ÷ 2.0606 = 113.6 mph.
Now run the question the other way. You want an 11.00-second car at the same weight. Rearranged, HP = W × (5.825 ÷ ET)3 = 3,500 × (5.825 ÷ 11.00)3 = 3,500 × 0.52953 = 3,500 × 0.1485 = 520 hp.
Read that carefully: one second of ET costs 120 horsepower, a 30% increase, on a car that already makes 400. Cube-root scaling is brutal that way, and it gets worse the quicker the car already is. Taking the same car from 11.00 to 10.00 seconds needs 3,500 × (5.825 ÷ 10.00)3 = 692 hp — another 172 hp for the same one second.
The cheaper lever is on the other side of the fraction. Because only the ratio matters, removing 350 lb from the 3,500 lb car has exactly the same effect as adding 40 hp to the 400: both take the ratio to 7.875 lb/hp and the ET to 11.59 seconds. Weight is usually the less expensive tenth.
How to read the prediction against your real timeslip
Compare the two predictions separately, because they diagnose different things.
If your trap speed matches but the ET is slow, you are losing the run at the start. Look at the 60-ft time on the slip: a full-bodied car on drag radials that leaves well is generally in the 1.6–1.8 second range, and every tenth you give away there costs you roughly a tenth and a half by the stripe. The fix is suspension, tyre pressure, launch rpm and converter, not power. The torque converter slip calculator and the shift point rpm calculator address the two most common culprits.
If the trap speed is also down, the power number you entered is wrong. Chassis dynos read wheel power, and quoting a wheel figure as flywheel power will make this calculator predict a car that does not exist. Convert it properly with the drivetrain loss calculator first.
If both are quicker than predicted, congratulations — you either have more power than the dyno said or the car is unusually slippery. A low-drag body genuinely beats the correlation at the top end, because the constants were fitted to cars with the aerodynamics of a brick.
One more correction matters before you compare anything: air. These correlations assume standard air, and a hot, humid, high-elevation day can cost a naturally aspirated engine several percent of its power. Run your track conditions through the density altitude calculator before you conclude the car is broken.
ET and trap speed by weight-per-horsepower (Fox constants)
| lb per hp | Cube root | ET (s) | Trap (mph) | Example car |
|---|---|---|---|---|
| 4.0 | 1.5874 | 9.25 | 147.4 | 3,200 lb / 800 hp |
| 5.0 | 1.7100 | 9.96 | 136.8 | 3,250 lb / 650 hp |
| 6.0 | 1.8171 | 10.58 | 128.8 | 3,600 lb / 600 hp |
| 7.0 | 1.9129 | 11.14 | 122.3 | 3,500 lb / 500 hp |
| 8.0 | 2.0000 | 11.65 | 117.0 | 3,200 lb / 400 hp |
| 9.0 | 2.0801 | 12.12 | 112.5 | 3,600 lb / 400 hp |
| 10.0 | 2.1544 | 12.55 | 108.6 | 3,500 lb / 350 hp |
| 12.0 | 2.2894 | 13.34 | 102.2 | 3,600 lb / 300 hp |
| 15.0 | 2.4662 | 14.37 | 94.9 | 3,000 lb / 200 hp |
| 20.0 | 2.7144 | 15.81 | 86.2 | 3,000 lb / 150 hp |
Every row satisfies ET × MPH ≈ 1,363, which is simply 5.825 × 234. Use that product as a sanity check on any prediction.
Safety thresholds come with the ET, not with the intent
Sanctioning bodies gate safety equipment on how quickly the car goes, not on what class you enter. The NHRA Rulebook has long set thresholds around the 11.50-second and 10.00-second marks for roll bars, roll cages and competition licences, with a separate speed threshold near 135 mph. Those numbers are revised, so treat the warnings this calculator raises as a prompt to read the current NHRA Rulebook and talk to your track's tech inspector — not as the rule itself. A car that predicts 10.4 seconds today is one good air day and one tenth away from being turned back at the gate.
Mistakes that make the prediction wrong
- Entering wheel horsepower. The constants were fitted against flywheel power. Feeding a chassis-dyno number in makes the car look 15–25% weaker than the correlation expects, depending on the drivetrain.
- Using curb weight. Race weight includes the driver, the fuel, the tools you forgot in the boot and the spare you never took out. A 180 lb driver alone is worth about 0.2 seconds on a 3,500 lb car.
- Assuming the constants cover a bad launch. They do not. They describe a car that hooks. That is what the traction setting in this calculator adjusts for.
- Applying it to a heavily boosted car with a huge power band. The constant-power assumption behind the cube root fails when power arrives in a step, and turbo cars often trap far above what their ET suggests.
- Ignoring the air. A prediction is a standard-air prediction. At 4,000 ft of density altitude a naturally aspirated engine is meaningfully down on power, and the ET moves with it.
- Comparing a Fox prediction with a Huntington one. They differ by nearly a second on a fast car. Pick one and stay with it.
Where these correlations sit among the alternatives
A cube-root correlation is the cheapest useful model of a drag pass. Above it sit two better tools and below it sits one worse one.
Better: a full simulation. Commercial drag-racing simulators integrate the equations of motion in small time steps using your actual torque curve, gear ratios, tyre radius, frontal area and drag coefficient, and a traction model. They predict incremental times as well as the final numbers and they respond correctly to a gear change. If you are chasing hundredths, that is the tool.
Better for diagnosis: your own data. Two runs in similar air with one change between them beats any model. Correlations tell you what a car like yours should do; the timeslip tells you what yours does.
Worse: power-to-weight rules of thumb without the cube root. "Ten pounds per horsepower is a twelve-second car" is roughly right at that specific ratio and wrong everywhere else, because the relationship is not linear. Use the power-to-weight ratio calculator for the ratio itself, then bring it here for the time.
If you race an eighth-mile track, do not apply these constants directly to your slip. Convert the numbers first with the eighth to quarter mile conversion calculator, and remember that the conversion factors are themselves empirical fits stacked on top of these ones. Errors compound.
Terms on a timeslip
- Elapsed time (ET)
- Time from the moment the car rolls out of the stage beam to the moment it crosses the finish line 1,320 ft away. Reaction time is not part of it.
- Trap speed
- Average speed over a speed trap ending at the finish line, on most tracks the final 66 ft. It is the best single indicator of power-to-weight on a timeslip.
- 60-ft time
- Elapsed time over the first 60 ft. The single most sensitive measure of how well the car left the line, and the number that most often explains a slow ET.
- Race weight
- The weight of the car with driver, fuel and everything else aboard, as it would cross the scales after a run.
- Flywheel horsepower
- Power at the crankshaft, before the drivetrain takes its cut. Manufacturer ratings and engine-dyno figures are flywheel numbers; chassis-dyno figures are not.