Engine power and power at the PTO are not the same number
A tractor’s engine makes one amount of power. What reaches the power take-off, after the transmission and everything else the engine has to turn, is a smaller amount. Makers usually quote the first. Work is done with the second. The two numbers answer different questions, and a specification sheet that gives you one of them has not given you the other.
What is the difference between engine power and PTO power?
They are measurements of different things, taken in different places.
| Figure | Measured where | What it tells you |
|---|---|---|
| Gross engine power | At the flywheel, often with the fan, alternator and exhaust removed or unloaded | What the engine can make in the best case. Not what the machine delivers |
| Net engine power | At the flywheel, with the engine’s own accessories fitted | Closer to reality, still before the transmission |
| PTO power | At the power take-off shaft | What a mounted implement can actually take |
| Drawbar power | At the drawbar, pulling | What a trailed implement can take. Lower again, and depends on ballast, tyres and the surface |
Each of these is a genuine measurement. None of them is wrong. They simply are not interchangeable, and swapping one for another is the commonest way a specification comparison goes wrong.
How big is the gap, really?
We can only answer that for machines where we hold both kinds of figure: a power claim from the maker, and alaboratory-measured PTO figure from a standardised test. That is 185 of the 3,122 tractors on the site. On those:
- The middle machine’s claimed figure sits 4.7% above its measured PTO power.
- Half of them fall between 0.6% and 8.6%.
- 145 of 185 have a claimed figure above the measured one, as you would expect.
- But 40 of them — about 22% — run the other way: the maker’s claim is lower than the power actually measured at the shaft.
That last line is the important one, and it is why this page does not give you a rule of thumb to multiply by.
Why is a claimed figure sometimes lower than the measured one?
Because on most of these machines the maker never said which standard the claim came from. A figure printed as “horsepower” with nothing after it might be gross engine power, net engine power, or power at the PTO filed under the engine’s name — and those are three different quantities. When the claim happens to be a conservative net or PTO number and the test day went well, the measured figure comes out on top. Nothing has gone wrong; the comparison was never between like and like in the first place.
This is exactly why we store every power figure with its standard and its basis, and never collapse them into a single “hp” column. Where a figure’s standard is not stated, we say so on the page rather than quietly treating it as gross.
Why does one tractor have several horsepower figures?
Because one laboratory test produces more than one number. A single OECD or Nebraska run measures power at the shaft at several conditions — at the speed where the engine makes the most power, at the speed the maker rates the engine at, and at the standard shaft speed of 540 or 1,000 rpm — and reports all of them. They are not attempts at the same figure. They are different questions, each answered once.
189 tractors on this site hold more than one measured PTO figure. Take the highest and the lowest of each without asking which run they came from, and they look as though they disagree by a middling 7.8%, and on one machine by 21.3%. That is where “the same tractor tests differently” comes from, and it is an artefact of reading a table wrongly.
Line the figures up properly — one machine, one tested version, one condition, one report — and across96 such groups the spread is a median of 0% and never more than 0.9%. That is the repeatability of the test itself, and it is very good.
The whole of the apparent disagreement is accounted for, and each part of it is a real thing:
- The rated-speed run is lower than the maximum-power run — by a median of 6.6% across 227 pairs from the same report. The engine does not make its most power at the speed it is rated at.
- The standard-shaft run is lower again — by a median of 2.4% across 227 pairs. Holding the shaft at exactly 540 or 1,000 rpm fixes the engine speed, and that is rarely the speed where the power is.
- Two versions of one model are two machines — on 33 models the laboratory tested more than one version, and their maximum-power figures differ by a median of 0.2% and by as much as 11%. A different gearbox or a different drive is a different result, honestly.
- And the maker's claim is not any of them, because it is engine power rather than shaft power — the section above.
This is why every figure on this site carries the run it came from. A PTO figure without its condition is not a worse number; it is a number you cannot use, because you cannot tell what it would have to be compared with.
What should I do with this when I am comparing tractors?
- Compare like with like. A measured PTO figure against another measured PTO figure. Two claims against each other, knowing both may be measured differently.
- Ask which standard a number came from. “110 hp” is not an answer; “81 kW at the PTO, OECD test” is.
- Treat a small difference as no difference. Two claims a few per cent apart tell you nothing: they can come from different standards, different eras and a conversion that rounded on the way.
- For mounted implements, the PTO figure is the one that matters. For heavy trailed work, look at drawbar power and remember it moves with ballast and ground conditions.
What does WhichTractor publish?
Both figures, labelled, wherever we hold them. Every power figure on a model page carries its standard and whether it is a claim or a measurement, and the two are never merged. 189 of the 3,122 tractors on the site carry at least one laboratory-measured PTO figure; the rest carry the maker’s claim, marked as a claim. Where the standard behind a claim was never stated, the page says that too — because a number whose provenance is unknown is worth less than one that admits it.