A study was performed to determine the effects of engine operating variables and piston and ring parameters on the crevice hydrocarbon emissions from a spark-ignition engine. Natural gas was used as the test fuel in an effort to isolate crevice mechanisms as the only major source of unburned hydrocarbons in the test engine's exhaust. The largest of the in-cylinder crevices, the piston ring pack crevices, were modified, both in size and accessibility, by altering the piston top land height and the number of piston rings and their end gaps. Each piston and ring configuration was subjected to a series of test sweeps of engine operating variables known to affect exhaust hydrocarbon emissions.None of the physical crevice modifications had any significant effect on the level of the exhaust hydrocarbon emissions, although the cycle-to-cycle repeatability of these emissions, measured with a fast hydrocarbon analyzer, was found to vary between the different configurations. Exhaust hydrocarbon concentrations were minimized by engine operation at leaner than stoichiometric fuel/air equivalence ratios and, at a fixed ratio, by choosing engine operating parameters which maximized exhaust temperature.A linear correlation between the exhaust hydrocarbon and oxygen concentrations was found at a stoichiometric fuel/air equivalence ratio. Crevice storage and release of unburned fuel/air mixture was thought to be the reason for this interrelationship.A correlation was derived between the exhaust hydrocarbon concentration and the exhaust temperature and engine speed which provided an excellent fit to the experimental data. It was concluded that, for an engine fuelled with natural gas, at a fixed fuel/air equivalence ratio, exhaust hydrocarbon levels are primarily determined by the degree of post-combustion oxidation as dictated by the post-combustion in-cylinder temperatures andresidence times.