The Motor Octane Number (MON) ranks fuels by their chemical resistance to knock. Evaporative cooling coupled with fuel chemistry determine Research Octane Number (RON) antiknock ratings. It is shown in this study that fuel Octane sensitivity (numerically RON minus MON) is linked to an important difference between the two test methods; the RON test allows each fuel's evaporative cooling characteristics to affect gas temperature, while the MON test generally eliminates this effect by pre-evaporation.In order to establish RON test charge temperatures, a computer model of fuel evaporation was adapted to Octane Engine conditions, and simulations were compared with real Octane Test Engine measurements including droplet and gas temperatures. A novel gas temperature probe yielded data that corresponded well with model predictions. Tests spanned single component fuels and blends of isomers, n-paraffins, aromatics and alcohols. Commercially available automotive and aviation gasolines were also tested. A good correlation was observed between the computer predictions and measured temperature data across the range of pure fuels and blends.A numerical method to estimate the effect of precombustion temperature differences on Octane sensitivity was developed and applied to analyze these data, and was found to predict the widely disparate sensitivities of the tested fuels with accuracy. Data are presented showing mixture temperature histories of various tested fuels, and consequent sensitivity predictions.It is concluded that a fuel's thermal-evaporative behavior gives rise to fuel Octane sensitivity as measured by differences between the RON and MON tests. This is demonstrated by the success, over a wide range of fuels, of the sensitivity predictor method described. Evaporative cooling, must therefore be regarded as an important parameter affecting the general road performance of automobiles.