Engine Horsepower Calculator

What an engine horsepower calculator measures

Horsepower is a rate of doing work. One mechanical horsepower is 550 foot-pounds per second, which is 745.6999 watts (NIST Guide to the SI, Appendix B.9). An engine that makes 300 hp is doing the work of lifting 550 pounds 300 feet every second, at least at the crankshaft on a test bench.

Everything an engine horsepower calculator does starts from that definition plus two things you can measure or look up: how hard each power stroke pushes (torque) and how fast the pushes arrive (rpm). There is no separate property called horsepower hiding inside an engine. It is torque delivered over speed, and any route to it is an estimate of that product.

This page works from the crankshaft. Wheel horsepower is a different number, lower by the share the drivetrain eats, and it has its own treatment. If you have a dyno sheet from the wheels, convert it back to crank power first and then compare.

Two more labels show up on power figures. Indicated horsepower is what the engine develops inside the cylinder, measured or inferred before the friction and pumping losses, and it is always higher than what reaches the dyno. Brake horsepower is what is measured at the crankshaft with the engine turning, and it is the number a dyno sheet reports. Since a good engine loses 10 to 15 percent of its indicated power to friction, pumping and windage, a claim of 350 indicated hp points to about 300 brake hp.

Torque, rpm and the 5,252 constant

Torque in pound-feet already is work per revolution. Horsepower turns that into work per second, and one horsepower is 550 foot-pounds per second. One revolution at a given torque moves the equivalent of 2 pi times the torque in foot-pounds, so:

hp = torque (lb-ft) x rpm / 5,252

The 5,252 is not arbitrary: 550 ft-lb per second x 60 seconds = 33,000 ft-lb per minute, and one revolution turns that through 2 pi radians, so 550 x 60 / 6.2832 = 5,252. Below 5,252 rpm an engine makes more torque than horsepower; above it, more horsepower than torque. Every dyno plot crosses over at exactly 5,252 rpm for that reason.

Read the table left to right and you can see why high rpm engines make big numbers from small torque. Double the rpm at the same torque and horsepower doubles, even though the engine never pushes harder.

The formula also explains why engine builders talk about a torque peak and a power peak separately. Torque is the force in the cylinder, and it depends on how much air is burning and how efficiently. Horsepower adds the speed of the fire. An engine can make the most torque it will ever make at 2,500 rpm and still have most of its power left at 6,000, which is why a wide-open drag race is won on the top end and not on the peak torque number.

Horsepower from displacement and BMEP

Before an engine exists, engineers estimate its power from brake mean effective pressure (BMEP), the average pressure pushing the pistons down through the power stroke. For a four-stroke engine:

hp = BMEP (bar) x displacement (L) x rpm / 895

Divide both sides by the displacement and you get hp per liter = BMEP x rpm / 895, the number engine builders use to compare one design against another. Typical BMEP bands are narrow, which is what makes the estimate useful.

A diesel makes less power per liter than its 22 bar pressure suggests because it turns so slowly: pressure times displacement times rpm, and a low rpm caps the third factor no matter how hard the first one works.

Cylinder count and stroke type change the constants, not the method. A four-stroke engine has one power stroke every two revolutions, which is where the 1,200 in the kilowatts formula comes from. A two-stroke fires every revolution and makes about twice the power per liter, so a 125 cc motocross two-stroke at 38 hp works out to 304 hp per liter, far outside every band in the table.

Worked example: a 2.0 L turbo four

Take a 2.0 L turbocharged four with the torque curve below, measured on an engine dyno with the SAE correction applied. Horsepower at each point is torque x rpm / 5,252.

Peak torque of 260 lb-ft arrives at 3,500 rpm, but peak horsepower of 235 hp does not arrive until 6,500 rpm, because horsepower keeps climbing as long as rpm rises faster than torque falls. Torque at the power peak is only 190 lb-ft, less than three quarters of the headline torque figure, and that is normal.

The BMEP check closes the loop: 235 hp from 2.0 L at 6,500 rpm implies 235 x 895 / (2.0 x 6,500) = 16.2 bar, which sits inside the turbocharged mainstream band in the table above. If a claimed figure implied 40 bar, the claim, not the formula, is wrong.

Turbocharging shows up in the same arithmetic as added BMEP. A 2.0 L naturally aspirated engine at 11 bar and 6,500 rpm makes 11 x 2.0 x 6,500 / 895 = 159.8 hp. The same block at 18 bar makes 261.5 hp at the same speed, and the extra 102 hp came from air the engine could not previously swallow.

Gross versus net: the 1972 rating change

US automakers rated engines by SAE gross horsepower until the early 1970s, measured with no air cleaner, no muffler, no alternator and no cooling fan attached. Net ratings, adopted as the SAE J1349 procedure took hold and applied across the industry by the 1972 model year, install the engine as the car does. The same hardware loses 15 to 25 percent of its rated number in the changeover.

A 1970 advertisement listing 350 gross hp points to roughly 280 net hp at a 20 percent gap: 350 x 0.8 = 280. Classic car listings still mix the two standards, so the first question to ask about any pre-1972 horsepower figure is which standard it was rated under. Gross ratings did not measure a different engine; they measured a different installation.

The change also explains why classic horsepower numbers feel inflated. A 1970 muscle car advertised 350 gross hp and ran 13.5 seconds in the quarter mile, and those two numbers are both true at the same time. The gross rating was measured without the equipment the car carried, and the quarter mile number included all of it.

Altitude, weather and dyno correction

A naturally aspirated engine is an air pump, and thin air shrinks the pump. The usual rule of thumb is a loss of about 3 percent of power per 1,000 feet of elevation. A 300 hp engine at 5,000 feet loses about 15 percent: 300 x 0.85 = 255 hp. Denver driving costs a naturally aspirated engine one car length of acceleration out of every seven.

A turbocharged engine loses a fraction of that, because the turbo spins harder until it reaches its pressure ceiling. High heat and humidity also cost power, since hot humid air is less dense than the cool dry air the ratings assume.

Dyno numbers are corrected for this. The SAE J1349 standard corrects measured power to a reference day of 77 F and 99 kPa of dry air, so a correction factor accompanies every legitimate dyno sheet. An uncorrected number measured on a cold morning can read several percent above the same engine tested the same day in July, which is worth asking about before you compare two sheets.

Air temperature matters for the same reason altitude does. Density falls about 1 percent for every 10 F of temperature, so a 90 F day takes about 2 percent off a naturally aspirated engine compared with a 70 F day, and a cold dense morning at 40 F is worth a little more than the reference day the ratings assume.

Sanity checks: hp per liter and fleet averages

When a claimed figure looks wrong, compare its specific output against real engines.

Fleet data backs the same bands. According to the 2025 edition of EPA's Automotive Trends report, the average new vehicle produced 258 hp in model year 2024, and the specific power of gasoline engines has climbed about 0.02 horsepower per cubic inch every year for 50 years (EPA Automotive Trends data). Displacement fell over the same period, which is the whole story of turbos, direct injection and higher compression.

Outside those bands, do not assume the engine is exotic. Assume the rating is measured on different equipment, at the wheels instead of the crank, or measured wrong.

Weight completes the picture, because horsepower only means something next to the mass it moves. EPA tracks horsepower and weight together, and the average new vehicle in model year 2024 carried 4,354 pounds and 258 hp, which is about 16.9 pounds per horsepower. A 1985 compact with 90 hp and 2,300 pounds was at 25.6 pounds per hp, which is why it felt lively in a different way from a modern car with the same output.

How to check a horsepower claim

Three checks catch most inflated claims without a dyno.

Convert the units. European spec sheets quote kilowatts: 1 kW is 1.341 hp, so a 300 kW engine is 402 hp, not 300. Metric horsepower (PS) is a third unit, worth 98.6 percent of a mechanical horsepower.

Back out the BMEP. Divide claimed hp by displacement and rpm using the formula above. Anything above about 135 hp per liter on pump gasoline needs racing fuel, high boost or an explanation.

Ask what the number includes. Crank or wheels, corrected or not, gross or net, and at what rpm. A figure quoted without an rpm is close to meaningless, because the same engine produces different horsepower at every speed on the curve.

For engine swaps and builds, run the calculator in the BMEP mode first to size the hardware, then compare the projected torque curve against the transmission and rear axle ratios you plan to use. The calculator is an estimator, not a substitute for the manufacturer's tested rating.