In-cylinder heat flux, cylinder pressure, and flame arrival and position data were obtained at air fuel ratios ranging from 11 - 16 at 3060 rpm and approximately 80% load. The engine used was a single cylinder, 5 hp, fixed timing, four stroke, overhead valve, air-cooled engine. Methods of mixture preparation include that produced with the stock carburetor, and with a system designed to provide the engine with a homogeneous mixture (HMS). Heat flux was measured using a thermopile device consisting of 300 thermocouple pairs. A thin film platinum RTD was used to measure the temperature at the thermopile and correct for sensitivity of the thermopile output to thermopile temperature. Flame arrival near the sensor was found through the analysis of an ion voltage signal from a probe located next to the heat sensor.An effort was made to identify and account for the variables which influence in-cylinder heat transfer. It was determined that with changing A/F mixtures, the dominant factor influencing heat transfer was variation in flame and pressure development which led to changes in both peak heat flux and integrated heat flux. Flame temperature variation with mixture A/F and amount of cyclic variation has a moderate affect on in-cylinder heat transfer. Amount of fuel vaporization, and mixture specific heat variations with A/F do not appear to have much effect.Variations in average peak heat flux, location of peak heat flux, and integrated heat flux were investigated over the range of mixtures. It was found that average peak heat flux was a maximum at the peak power mixture (A/F=13.0) while the integrated heat flux was at a maximum closer to the stoichiometric mixture (14.5). Integration of the instantaneous heat flux over the entire cycle was broken down into a compression and expansion component and a gas exchange component. For compression and expansion, the integrated flux peaks at the peak power mixture and varies with A/F similarly to the average peak heat flux curve. In contrast, for gas exchange, the integrated heat flux peaks just lean of stoichiometric due to increased exhaust gas temperatures. The higher exhaust gas temperatures at these leaner mixtures is primarily a result of slower burn rates.Correlations were obtained between the phasing of peak pressure, phasing of peak heat flux, and flame arrival. Correlation coefficients greater than 0.99 were obtained for a linear fit between average phasing of peak heat flux and average phasing of peak pressure. Another correlation investigated was between flame arrival time and time of peak heat flux. In this case, correlation coefficients greater than 0.96 were obtained.