There are a lot of misunderstandings among car enthusiasts and historians about vintage horsepower ratings. It’s easy to assume from a casual glance at ads or spec sheets that even quite ordinary American family sedans of the sixties were overwhelmingly powerful, with 300 horsepower or more, and yet by 1975, many of those same cars were down to 150 hp or less. When asked the reason for the huge difference, gearheads tend to shake their heads and mutter about emissions controls and anemic, low-octane unleaded gasoline — which is true, but only partly.
What complicates the issue and makes apples-to-apples comparisons difficult is the fact that those pre-smog horsepower ratings were not calculated in the same way as modern engines. “A horsepower is a horsepower, right?” you say. While a horsepower, pre-smog or post, remains 746 watts (or 736, for metric horsepower), the way that output was measured has changed quite a bit. Let’s explain:
GROSS HORSEPOWER RATINGS
Over the years, the Society of Automotive Engineers (SAE) and similar standards-setting bodies in other countries have developed various methodologies for measuring the output of an automotive engine. These standards fall into two basic categories: gross and net output.
Gross output, which in the U.S. is typically measured using the methodology laid out in SAE standards J245 and J1996, is the output of a ‘bare’ engine running on a test stand with no accessories, free-flowing exhaust headers (no mufflers), and optimal ignition timing. It also includes a correction factor for standard atmospheric conditions. In other words, gross output represents a particular engine’s maximum output under ideal conditions.
In the real world, automotive engines very rarely operate in ideal conditions. The engines of most cars are burdened with various engine-driven accessories, including the engine’s own oil pump, water pump, and generator/alternator and add-ons like the power steering pump and air conditioning compressor. Engines intended for on-road use typically also have restrictive air cleaners and exhaust systems, sound-deadening mufflers, and emissions-control add-ons like catalytic converters and thermal reactors. Engine tuning is further compromised in the interests of reduced noise, better drivability, improved cold-weather performance, and lower exhaust emissions. All of these factors reduce the engine’s maximum output in ways that the gross rating methodology does not reflect.
For that reason, the SAE and similar bodies have also established standards for measuring net output. Net ratings, such as the ones defined by SAE standards J1349 and J2723, are still taken with the engine on a test stand, but reflect stock ignition timing, carburetion/fuel delivery, exhaust systems, and accessories. The specific methodology varies depending on the standard applied, but the gist is that a net rating is a closer approximation of an engine’s output as actually installed in a car.
Naturally, the net output of a given engine is somewhat lower than the gross output. For example, the 217 cu. in. (3,547 cc) “Stovebolt” six in a 1950 Chevrolet had a gross output of 92 hp (69 kW), but a net output of only 85 hp (63 kW). Chevrolet’s 1955-vintage 265 cu. in. (4,344 cc) small block V8, meanwhile, had a gross output of 162 hp (121 kW) with 8.0 compression and a two-barrel carburetor, but a net output in the same form of 137 hp (102 kW).
ADVERTISING, OVERRATING, AND UNDERRATING
Net ratings have existed for many years, but before 1972, most U.S. automakers preferred to quote the more impressive-sounding gross figures. Net figures might appear in shop manuals, technical papers, or other factory literature, but rarely showed up in advertising.
Until the mid-fifties, the gap between gross horsepower and as-installed output was seldom vast. However, as it became apparent that impressive horsepower numbers sold more cars, manufacturers’ advertised gross output figures began to climb. While automakers didn’t necessarily publish net ratings for their hot engines, the performance of cars so equipped suggested that the advertised figures were now substantially higher than the engines’ as-installed outputs — sometimes by 25% or more.
The late-fifties recession temporarily put the brakes on the advertised horsepower race. By the mid-sixties, the advertised outputs of many bread-and-butter engines were still unrealistically high, but the outputs of Detroit’s most powerful engines were now as likely as not to be deliberately understated. For example, in 1965, Chevrolet released the 396 cu. in. (6,488 cc) Turbo-Jet V8 as an option for Corvettes, rated at 425 gross horsepower (317 kW). The following year, the engine was bored to 427 cubic inches (6,996 cc). At first, some Chevrolet promotional material quoted a gross output of 450 hp (336 kW) for the bigger engine, but the division then hastily restated maximum power as 425 hp (317 kW), no more than the 396. Outside observers were incredulous, particularly since the bigger engine was plainly more powerful than the smaller version.
Why would a manufacturer underrate their engines? One possible reason was racing. Particularly in drag racing, eligibility for different classes was sometimes based on advertised horsepower and shipping weight, so an engine that produced more power than its advertised gross rating indicated offered an obvious advantage. Racing officials were seldom fooled by such tricks for long, however, and began to handicap or “factor” advertised outputs to place a conservatively rated engine in a more appropriate class.
Another reason for underrating was political decorum. Particularly at GM, the most conservative of the Detroit automakers, there was real fear of the safety lobby, which already considered the power outputs of existing engines to be irresponsibly high. In that climate, offering a 450 hp (336 kW) or 500 hp (373 kW) Corvette, for instance, would have been asking for trouble. Insurance was also becoming an issue, with a growing number of insurance companies levying prohibitive surcharges on very powerful cars or simply refusing to cover such cars at all.
For those reasons, GM also had internal rules limiting all of their passenger cars except the Corvette to a maximum of one gross horsepower per 10 pounds (1 kW per 6.1 kg) of curb weight, leading to curious non sequiters like rating Pontiac’s 3,300 lb (1,500 kg) Firebird at 325 hp (242 kW) while claiming 360 hp (269 kW) for the mostly identical engine in a 3,600 lb (1,635 kg) GTO.
Underrating of this kind was at best an open secret. When Car Life magazine tested a 1969 Pontiac GTO Judge equipped with the $390 Ram Air IV engine, for example, Pontiac executives freely admitted the engine’s 370 hp (276 kW) gross rating was purely a political fiction.
NET HORSEPOWER RATINGS
Between inflation and deliberate underrating, the relationship between advertised gross horsepower and actual net output was becoming increasingly nebulous by 1970. The gross ratings served a variety of useful marketing and political purposes, but as a realistic measure of engine output, they left much to be desired. In particular, the gross ratings seldom reflected the impact of emissions control devices like air injection and exhaust gas recirculation, which were already beginning to have a noticeable impact on engine performance.
Faced with that reality, along with the pressures of the safety and environmental lobby, U.S. manufacturers decided around 1971–1972 that it was time to abandon the gross rating system in favor of more conservative SAE net ratings.
The immediate result was a dramatic drop in advertised power. For example, the mammoth 500 cu. in. (8,194 cc) engine in the Cadillac Eldorado fell from 400 gross horsepower (298 kW) in 1970 to only 235 net horsepower (175 kW) for 1971. The real decline wasn’t quite as steep as it looked; 1971 engine did have a lower compression ratio to prepare for the adoption of unleaded gasoline, but the 1971 gross rating was still 360 hp (269 kW), so the actual loss was about 10%, not more than 40%. (Cadillac did not publish net horsepower ratings for the high-compression 1970 engine, but our guess would be that it made 275–285 hp (205–213 kW).)
Why was this change made? Given the pressure of the new emissions standards, the switch to a more realistic rating system was timely and probably prudent; consumers can and periodically do sue manufacturers whose products don’t perform as advertised. The move also provided a useful and relatively inexpensive PR gesture. By simply switching from gross to net ratings, automakers sent a message to lobbyists and lawmakers that the horsepower race was over and Detroit was no longer offering outrageously powerful engines. Beyond that, the timing of the switch helped to obfuscate the actual losses caused by added emission control hardware and lower compression ratios, which was presumably very useful for the unfortunate Cadillac salesman trying to rationalize why the 1972 Cadillac the customer is looking at seems to have 40% less power than the 1970 model he’s trading in.
Rating systems aside, the initial decline in power was relatively modest, but that wouldn’t remain the case for long. For example, the hottest version of Pontiac’s 455 cu. in. (7,481 cc) V8 still managed 310 net horsepower (231 kW) in 1973, but was down to only 200 hp (149 kW) by the time it faded out in 1976. Ford’s familiar 302 cu. in. (4,942 cc) V8, meanwhile, which in the sixties had advertised as much as 306 gross horsepower (228 kW), had plummeted by 1979 to less than 140 net horsepower (104 kW). It was not until the widespread proliferation of electronic fuel injection in the 1980s that net power outputs again began to climb.
U.S. automakers continue to use SAE net ratings, but in 2005, the SAE issued a new standard, J2723, which clarified and revised the existing methodology, among other things requiring the presence of an independent observer. Some engines measured under the new “SAE-certified output” guidelines ended up with lower ratings than before while a few other actually ended up with higher ratings. In most cases, the engines were not actually altered in any significant way; the changes were in the test methodology. Most if not all manufacturers now use this methodology for the U.S.-market cars and trucks.
NON-U.S. HORSEPOWER RATINGS
What about cars not built by U.S. automakers? As with many things, the answer is, “It depends.”
German automakers have long rated their engines under the DIN (Deustches Institut für Normung, German Institute for Standardization) standards. DIN horsepower and torque figures are metric net ratings, similar but not identical to SAE net ratings due to minor differences in test methodology. Some Italian automakers formerly used the Comitato unitario dell’autotrasporto (CUNA) standards, whose metric net ratings again were similar but not identical to SAE net ratings. Depending on the time period, model, and market, other European automakers typically used either DIN or SAE gross figures.
Until the early 1970s, some British automakers quoted SAE gross figures for horsepower and torque while others listed net figures; some manufacturers listed both. After the U.K. joined the European Economic Community in 1972, the British auto industry switched to DIN ratings.
Japanese automakers rate the outputs of their home-market cars under the Japanese Industrial Standards (JIS), which include methodologies for both gross and net output. The difference between the two was generally around 15%, sometimes a little more, sometimes a little less. Until the mid-eighties, most Japanese automakers quoted JIS gross figures for most if not all products sold in the domestic market. The switch to JIS net ratings began around 1985, but wasn’t completed until late in the decade; confusingly, for some model years, manufacturers would quote net ratings for some engines and gross ratings for others even within a single model line.
Prior to 1972, most non-U.S. automakers would publish SAE gross figures for all engines exported to the U.S. However, foreign automakers seldom indulged in the kind of games Detroit sometimes played with its gross ratings, so the differences between gross and net ratings were typically small and probably fairly realistic. For example, the Triumph TR4 carried a gross rating of 105 hp (78 kW) and a net rating of 100 hp (75 kW) while a 1963 Mercedes-Benz 230SL had a gross rating of 170 hp (127 kW) SAE and a net rating of 150 PS (110 kW) DIN.
MECHANICAL VS. METRIC
There is an additional complication when considering non-U.S. power ratings: the question of units. DIN, JIS, and CUNA standards are typically — but not always — quoted in terms of metric horsepower rather than the mechanical horsepower more familiar to our American and British readers. One metric horsepower (often abbreviated PS) is about 736 watts while one mechanical horsepower is about 746 watts, so 1 PS equals 0.986 hp. For example, 300 metric horsepower would be about 296 mechanical horsepower while 150 PS is about 148 hp.
Unfortunately, many sources are maddeningly inconsistent in their application and even understanding of these units and tend to freely interchange them, always writing “hp” or “bhp” even when talking about PS. Occasionally, some sources will attempt to convert one unit to another, sometimes inconsistently — for example, multiplying by 0.986 a rating that was already in mechanical horsepower. (We’ve almost certainly been guilty of that, although we try to avoid it.) Since the difference is less than 2%, it’s not a big deal, but it does frustrate a lot of efforts to be consistent or precise.
SOME COMMON MISCONCEPTIONS
Let’s clear up a couple of common misconceptions about gross and net horsepower ratings:
- Contrary to some assumptions, net horsepower ratings do NOT measure horsepower at the drive wheels. Both gross and net ratings are at the flywheel and do not reflect power losses in the drivetrain.
- Because of the vagaries of the old gross ratings and the widespread over- and underrating of different engines, there is NO precise formula for converting gross horsepower to net horsepower or vice versa. Sometimes, the difference is as little as 5–10%; sometimes it’s more like 25%. Unless the factory released both gross and net figures for a given engine (which some did, even in the U.S.), the best you can do is make an educated guess based on state of tune and real-world performance testing — keeping in mind that published road tests didn’t necessarily reflect the performance of cars the average consumer could actually buy.
- The ambiguity of gross horsepower ratings means that many pre-1972 American cars were actually a lot less powerful than the advertised figures would suggest. While the late sixties were a golden age of horsepower compared to the late seventies or early eighties, the differences weren’t quite as vast as they appear at first blush. Some engines’ advertised gross horsepower probably exceeded their actual as-installed power by 100 hp (75 kW) or more.
NOTES ON SOURCES
Most of this information is based on decades of reading car magazines. Some of our specific references included (but were not limited to) Robert Ackerson, Chrysler 300 ‘America’s Most Powerful Car’ (Godmanstone, England: Veloce Publishing Plc., 1996); “Alfa Super Spider,” Road & Track October 1959, reprinted in Road & Track on Alfa Romeo 1949-1963, ed. R.M. Clarke (Cobham, England: Brooklands Books Ltd., ca. 1985), pp. 52–53, 65; the Auto Editors of Consumer Guide, Consumer Guide Car & Truck Test Monthly August 2005 and July 2006 and Encyclopedia of American Cars: Over 65 Years of Automotive History (Lincolnwood, IL: Publications International, 1996); John R. Bond, “Car of the Year: The 1949 Cadillac,” Motor Trend November 1949, reprinted in Cadillac Automobiles 1949-1959, ed. R.M. Clarke (Cobham, England: Brooklands Books Ltd., ca. 1985), pp. 8-9; Bernard Cahier, “Road Test/10-63: Mercedes-Benz 230 SL,” Sports Car Graphic May 1963, reprinted in Mercedes 230SL – 250SL – 280SL Ultimate Portfolio 1963-1971, pp. 31-34; “Chevrolet Camaro Z28,” Car and Driver May 1971, reprinted in Camaro Muscle Portfolio 1967-1973, ed. R.M. Clarke (Cobham, England: Brooklands Books Ltd., ca. 1992), pp. 118-121; Chevrolet Division of General Motors Corporation, Chevrolet 1950 Engineering Features: Passenger Cars (Detroit, MI: General Motors Corporation, 1949); “GM: Cadillac,” Motor Trend Buyers’ Guide 1971, reprinted in Cadillac Eldorado 1967-78 Performance Portfolio, ed. R.M. Clarke (Cobham, England: Brooklands Books Ltd., ca. 2000), pp. 80-83; “GM: Oldsmobile,” Motor Trend Buyers’ Guide 1971, reprinted Oldsmobile Muscle Portfolio 1964-1971, ed. R.M. Clarke (Cobham, England: Brooklands Books Ltd., ca. 1999), pp. 138-140; Tony Grey, “Olds 4-4-2,” Road Test May 1971, reprinted in Oldsmobile Muscle Portfolio 1964-1971, pp. 132-137; John Gunnell, ed., Standard Catalog of American Cars 1946-1975, rev. 4th ed. (Iola, WI: Krause Publications, 2002); Roger Huntington, “Chevrolet’s New V-8,” Motor Life December 1954, reprinted in Chevrolet 1955-1957, ed. R.M. Clarke (Cobham, England: Brooklands Books Ltd., ca. 1988), pp. 16-17, 31; Jikayousha [Private Car] Buyer’s Guide Spring ’87 Edition, February 1987; Steve Kelly, “Beware the Quiet Fish,” Hot Rod January 1971, reprinted in Plymouth 1964-1971: Muscle Portfolio, ed. R.M. Clarke (Cobham, England: Brooklands Books Ltd., ca. 2003), pp. 120-122; John Lamm, “King of the Hill: Cadillac Eldorado vs. Lincoln Continental Mark IV,” Motor Trend July 1972, reprinted in Cadillac Eldorado 1967-78 Performance Portfolio, pp. 96-103; Brian Long, Celica & Supra: The book of Toyota’s sports coupes (Dorchester, England: Veloce Publishing, 2007); “Road Testing Chrysler’s Power Flite,” Speed Age November 1953, pp. 58-62, reprinted in Chrysler Imperial Gold Portfolio 1951-1975, ed. R.M. Clarke (Cobham, England: Brooklands Books Ltd., ca. 2004): 20-23; “The Judge: Car Life Road Rest,” Car Life March 1969, reprinted in GTO Muscle Portfolio 1964-1974, ed. R.M. Clarke (Cobham, England: Brooklands Books Ltd., ca. 1998), pp. 91-95; “Triumph TR4 (Road Test No. 26/62), The Motor 11 July 1962, reprinted in Triumph TR4 – TR5 – TR250 1961-1968 (Brooklands Road Test Series), ed. R.M. Clarke (Cobham, England: Brooklands Books Ltd., ca. 1997), pp. 30-33; and Mark Wan’s AutoZine (www.autozine.org). We also verified a few details per the Wikipedia® entry on horsepower (en.wikipedia.org/wiki/Horsepower).