"The terrible forty-five."

American Honda’s great racing manager Gary Mathers used those words because the RC45 was four years late in development (work was stopped twice for various reasons) and tricky to ride. It finally arrived in 1994. Yet it did win Superbike championships – once in World Supers, 1997, by John Kocinski, and twice in US AMA Superbike—by Miguel DuHamel in 1995 and by Ben Bostrom in 1998.

In its early years, World Superbike was a dual displacement formula – 750 cc for four-cylinder engines, and 1000 cc for twins. The clear intention of the rules-makers (Steve McLaughlin and John Ulrich) was that the higher rpm capability of the fours would compensate for the greater displacement of the twins. RC45 would be the ultimate expression of the 750, producing something close to 190 hp at 17,000 rpm.

Review the ‘45’s lineage. As Honda management saw FIM Grand Prix racing become 100% two-stroke-dominated in 1975, the very senior Tadashi Kume indicated it was time to show what four-stroke engines could do. Honda’s master of 20,000+ rpm engine design, Shoichiro Irimajiri, was made head of design.

It would turn out that the V4 engine architecture would be the lasting legacy of this work—NR-Block’s oval-piston NR500 GP four-stroke. Engineers liked what they saw in the V4. First, with a 90-degree cylinder bank angle, it could be given primary balance without a balance shaft. Second, its compactness was attractive because its x, y, and z polar moments (resistance to being set into rotation – as when “flicking” to the lean angle in cornering) were most like those of a cannon ball, and least like those of a ladder. And third, with Honda’s focus on friction reduction, its crankshaft would have 3 or 4 main bearings to the in-line four’s 5 mains.

The NR isn’t our subject here. From it came Honda’s V4 Sabre and 750 Interceptor, the first bike whose excellent handling drove its sales success. Lighter aluminum-framed VFRs won AMA Supers titles in 1986, ’87, and ’88.

To win the new World Superbike race series starting in 1988, Honda designed the VFR-inspired RC30 as a “homologation special” - a means of making its advanced features legal in a production-based class. Homologation usually requires that a certain minimum number be produced. As you’d expect, the RC30 has become a favorite of collectors. How many spend their days in temperature- and humidity-controlled rooms?

Now for an anomaly. In contemplating pitting high-revving 750-4s against the two big 500 cc cylinders of a twin, you’d expect wide torque from the twins to smoke the high revolutions of the four.

It didn’t happen. When Massimo Bordi, designer of Ducati’s 851 90-degree Vee-Twin, spent time at Cosworth, designers of F1 engines, he came away with “Cosworth valve timing” – long valve durations that gave best power on top, but were weak in the middle, forcing Ducati superbike riders to adopt a corner-speed style that was less dependent on acceleration.

Honda, on the other hand, initially designed the RC30’s V4 engine with broad torque that began at 7500 rpm that first year (1988). That proved to be the tool in World Superbike – Fred Merkel took the first two championships - 1988 and ’89.

Ducati’s 851-based design was more fluid than solid, always changing, displacement increasing, always with strong cylinder-filling and efficient combustion, but also with inherited defects that would take time to fix along the way.

As Honda raised power, RC30’s torque threshold rose with it – 8500 for 1989, and soon, with mods to match Ducati’s gains, torque threshold rose to 10,500 (one important boost to top-end came from what they called “prolapsed intake” – machining the faces of the intake ports back into the heads, shortening intake length.

In 1990, Raymond Roche gave Ducati its first World Superbike title. Doug Polen took the title on Ducati in ’91 and ’92, then Scott Russell (On the in-line Kaw ZXR-750).

This wasn’t Honda’s plan. With RC30’s torque band narrowing, this became a clear illustration of what obsolescence means. It arrives when everything has been pushed to its functional best; an obsolete race bike is one that can no longer be improved.



Historical sources are not forthcoming on causes of the two work stoppages in RC45 development. Insiders hint the bike arrived four years later than planned. In those years Honda was catching up to Yamaha in 500 cc GP racing. When the decision to can the NR four-stroke was made (piston fatigue problems at 19,000 instead of the planned 23,000) Honda and Freddie Spencer went two-stroke GP racing in 1982 with Shin’ichi Miyakoshi’s strategy of kindness to tires. Arriving in a 130 hp race series with 108 hp from a stone-simple reed-valve Triple, this new combo won two GPs the first year and the 500 title the next.

While RC45 was having work interruptions, Honda were also developing the “Big Bang” close-firing order concept of 1992. It worked, as did other innovations. Starting in 1993, Mick Doohan won five consecutive 500 GP titles on Honda’s NSR500. I put this bit in so you’ll know RC45 had competition for R&D resources.



At the time when RC30 was winning its two World Supers titles, bikes with its chassis qualities were just coming under suspicion.

Designed when high chassis stiffness was the Holy Grail in GP racing (at first it increased stability), the RC30’s twin aluminum spars joined a cast steering-head of unparalleled rigidity.

Why even mention this? Because riders had reservations about RC30:

Freddie Spencer said turn-in was a big guess—a gamble instead of a judgment—because he couldn’t feel what the front tire was doing.

Turn-in a big guess? What does that remind us of? Every bike with too much front-end chassis stiffness, that’s what. Casey Stoner had to develop his “drop down on the tire edges and SPIN” technique to make the 800 Ducati steer. Miguel had to do something similar with RC45. Lesson; if the front won’t steer, try the rear.

And in ’94, when RC45 did appear, its early Large Project Leader, Tatsu Horiike said, “Excellent engine. Terrible Motorcycle.” At the races – problems with chronic understeer. On race tires a bike can lean so far that its actual suspension is seriously no longer pointed the way bumps act – up and down. So its performance disappears as lean angle increases. Older chassis made of bendy steel tubes did well in this condition, because their flexibility let the tires track over small bumps and pavement roughness with little air-time. But the stiffer the chassis, the less it can do this, until at super stiffness the tires skip from crest to crest, leading to understeer by way of air-time.

RC45’s engine was steps ahead of RC30. To make room for bigger valves to permit operation at higher revs, the bore was increased to 72 mm from 70, allowing bigger valves of 29 int/24.5 exh. Valve included angle was radically reduced from the 38 degrees of RC30 to 26 degrees, making a flatter combustion chamber and piston crown that retained intake motion well to speed combustion. Compression ratio was raised to 13.2. If you can make that work without detonation, compression boosts torque at all rpm.



The number of main bearings was reduced from 4 to 3 by moving the cam gear drive from crankshaft center to the right-hand crank end. Cams had been supported by rolling bearings which offer little radial damping. The new design provided plain cam bearings (just as Ducati would adopt for 2001).

The RC30 had 38 CV carbs but RC45 initially had single-pintle injectors controlled by a nose-mounted electronics box with four knobs on its left face. One explanation revealed that each knob controlled a zone of fueling – #1 for 0-25%, #2 for 25-50%, &c, just as a carb has its four zones – idle, slide cutaway, needle/needle-jet, and main jet. It’s not easy to find information about this, but at the end of the 1990s something very different is described – the “deflective multi-hole injector.” Its purpose was “to enhance fuel atomization performance.”

In the same time frame, other manufacturers had tried and discarded pintle-type injectors in favor of what we have today – multi-hole injectors. Each laser-drilled hole is around the diameter of a human hair – four thousandths of an inch. Their tiny jets break up into even tinier fuel droplets that evaporate quickly

At the 1996 Suzuka 8-hour, rider comments weren’t good:
“No feedback from front end.”
“Can’t tell how hard to flick it in”
“Feels like if you pull it in to {turn} front end will go away.”

Time for effective action. We know Honda had been experimenting with softer chassis, and that Yamaha in 1993 had found its stiffer-than-ever 500 GP chassis that year was too stiff – it skated and hopped. Others had actually sawn out chassis members in desperation attempts to make their bikes hook up.

Another fascinating aspect of RC45 was that, like RC30 before it, on teardown, its con-rod big-end bearing shells, in the words of Gary Mathers, “…always looked bad, like we’d taken it apart the very minute before it seized.”

Increased oil pressure did nothing—the multi-colored streaking continued.

Evidently there was a big RC45 push made for the ’97 World Supers season – too much time had passed winless. Crankcase evacuation was combined with low-pressure oil scraper rings to find four extra horsepower. A newly homologated dual fuel injector system was credited with a 10 hp gain, plus a fatter mid-range. And crank oiling was switched away from automotive style (oil for the crankpins bled from the main bearings) to F1 style – end feed. With end feed, “centrifugal force” works for you, but with main bearing bleeds you have to force the oil radially inward to reach the crankpin drillings. With end feed, the streaking disappeared and only 12 psi oil pressure was required.

Chassis hook-up rapidly improved as the new concept of lateral chassis flexibility was applied. John Kocinski won nine World Superbike races (he went well in rain, too) and the championship. How do you increase the flexibility of a homologated chassis? Back in the tube-frame days teams varied tube wall thickness. In tech, as long as OD looks good, it’s stock.

Planners at Honda now saw a “structural problem” with the World Superbike displacements of 4-cylinder 750s versus 1000 cc Twins. What had changed? Carl Fogarty had constantly complained of weak mid-range to his Ducati crew, and they had done something about it – shortened valve timing and increased valve lift.



It’s easy to make power – just keep the intake valves open later after bottom center. This allows intake air – rushing at its peak velocity – to keep right on rushing into the cylinders, even though the pistons have begun to rise on compression. Great peak power is the result. But in the mid-range, where intake velocity is lower, it’s less able to keep rushing into the cylinder. Instead, the rising piston stops it, and even reverses it, pumping some mixture back out of the cylinder. This is the classic cause of weak mid-range.

Just closing the intakes sooner can’t fix this. The flow volume lost to shorter timing has to be restored by increased valve lift. As the work succeeded Fogarty switched to complaining of too much mid-range.

Lifting valves higher in less time requires higher valve train accelerations., which increases stress on the cam drives. Eventually Ducati mechanics were changing cam belts every day, and the final expression of their radical Vee-Twin was designed with steel silent chain cam drives.

Once Ducati had mid-range, their twin became too much for a 750 V4. Rather than trying to push some kind of HyperRC45 to 20,000 rpm, Honda shifted their World Supers focus to designing a 1000 cc twin planned from the ground up as a racer. A few would have to be sold as part of homologation – maybe more than that if they caught on. RC51 and Colin Edwards won the class in the bike’s first year—2000.

Ducati were meanwhile finishing development of their all-new ‘Testastretta’ (narrow-head) engine, which with Troy Bayliss recovered the title in 2001. Honda and Edwards fought back with further development, taking the title back in 2002.

Was that the end for RC45? No. There had been pressure for some time to return to four-stroke grand prix racing. Under the new name “MotoGP” it began in 2002. According to legend, Mick Doohan, in conversation with Dorna’s Carmelo Ezpeleta, said of the new class, “What the hell—give ‘em a liter!” Dorna was the TV rights-holder for MotoGP.

The new class was set at 990 cc (couldn’t let anyone think they were copying WSBK’s 1000). Honda set to work developing RC211V, with a 5-cylinder engine. Here on my calculator it says five RC45 cylinders would total 937.5 cc – with a little help that could become a 990 V4. Honda increased the bore by one millimeter and stroke by 1.3 mm.

In that way, the last-minute frantic R&D that brought the 1997 World Supers title within Kocinski’s reach was not just discarded after that success. It lived on as a head start, giving Honda and Valentino Rossi the first two MotoGP world titles – in 2002 and 2003.

In 2003 it was Honda 1-2-3.