. Good link. This could become another good thought experiment such as some other topics here did. . http://www.cbr250.net/forum/cbr250-performance/7217-gearing-horsepowers-better-half-3.html . An engine torque curve and horsepower curve are really just two different ways of looking at the same thing. Power is just torque over time. People like to look at the torque curve to get an idea of what is going on below 5,000 rpm because it has better resolution on the graph due to the horsepower curve not yet having much multiplication from the low rpm. The power doesn't yet look like much and the torque looks much bigger but they are none the less showing the same information. And power is the real metric of how much work can be done. Torque can be changed any way you want with gearing but that never changes the power at the wheel because you are trading increased rear wheel torque for decreased wheel rpm. Increasing the rear wheel torque WILL make the bike accelerate quicker but when you get to the steady state top speed for any given load such as the wind on a flat top speed run or gravity on a big hill, the gear that will be fastest is the gear that puts the rpm at the power peak. Not the torque peak. And not the redline. Even a drag strip run which is all about acceleration and using the most torque multiplication, will be the quickest on a vehicle like the CBR250R when shifting the higher gears at an rpm band that is much lower than the redline and is centered on the power peak. As was generously explained in the other thread by DieselMaxPower . The OP's link is a little too simplistic when it comes to fuel efficiency. It is possible for an engine to take in the same amount of fuel and air (disregarding the much higher losses at high rpm) at 8,000 rpm as it is at 4,000 because of the throttle plate's restriction and resulting intake vacuum. The real reason for generally better fuel economy at lower rpm is from less frictional, reciprocating, and thermal losses than the higher rpm. And from being closer to the first torque peak. . The first torque peak usually shows where the engine is operating most efficiently as set up by the cam timing, intake runner tuning, ect versus linear piston speed losses, flame front travel and combustion gas pressure curves, ect,ect. A Brake Specific Fuel Consumption chart (Which plots 4 parameters; engine rpm along the bottom, the torque that is produced along the right, increasing engine load which can be roughly increasing throttle position up the left, and finally the topo values for how much fuel is being used versus how much power is produced.) is the only way to really know the engine's best efficiency range. They usually show the best efficiency at just off of full throttle on either side of the first torque peak. Which is the bad news for fuel economy. Most engines are way too powerful to operate at full throttle for more than a few seconds without going way too fast. So small engines with barely enough power, like our 250, are more often operated near their most efficient range than an engine with plenty of extra power. The only way to get the bigger engine into it's efficient range is to pulse with a high throttle opening and then glide with the clutch in, and then pulse again. But this an advanced hypermiling technique that is not for everyone. Better to stick to a CBR250R, PCX150 scooter, or new Ford Fiesta. . Thermal losses, pumping losses, friction losses and powerband tuning, among many other factors, all come into the design equation. Using the smallest combustion chamber and the fewest of them minimizes the area which can lose heat (wasted fuel) to the head. But too small of a bore reduces the force of any given combustion pressure because it offers less area to work on. So the tendency for engine design to have a square (equal bore/ stroke) design. Slightly longer stroke than bore will be more efficient as it will move the torque peak to a lower rpm because of the longer crank arm. Interestingly, a square single will have less combustion chamber area to pick up heat than a square twin of the same displacement. Multiply it out for yourself. It's cool. And so is the CBR250R's engine. I'm sure you have noticed that the engine and radiator don't throw off much heat compared to your other bikes. Air cooled engines can run cylinder head temps of 350F/ 175C which, compared to liquid cooled engines which are limited to 100C to prevent boiling, waste even less heat. . Heat out the exhaust is another waste. Which can be reduced with Atkinson cycle engines . http://en.wikipedia.org/wiki/Atkinson_cycle . (a fancy way of saying that the cam left the intake valve open too long.) which seek to have no pressure remaining in the cylinder by the time the piston reaches the bottom. And they also reduce pumping losses with lower intake vacuum. At the expense of power versus displacement which never looks good to bench racers reading spec sheets. . Honda motorcycles is fighting for better fuel fuel efficiency in the big bike west with the ultra long stroke CBR500 platform and ultra low rpm tuned NC700. Either one of which can almost match our already exemplary CBR250R at 68 mpgUS. BMW has always been a fuel efficiency leader. The Ninja300 is another sporty and all around good bike taking a step in the right direction with longer stroke and longer gearing. Some small 110cc air cooled bikes which are popular in India can sell for $1000 and get 150 mpgUS at 40 mph.