A divided-chamber, stratified-charge engine was operated at different speeds and loads, and with two prechamber nozzle sizes. Pressure was recorded and flame propagation was filmed through a transparent head. For each condition, several sets of data were taken and the experimental scatter identified. Pressure and flame propagation were then computed with a two-dimensional model and compared with the measured ones. Several quantities which influence the computed results were not directly measured in the experiment and assumptions had to be made about them in the model. They include amount and state of cylinder air and residual products, turbulence, and wall heat transfer. A turbulent diffusivity, an overall reaction rate, and a wall heat loss proportional to the heat released thus were adopted in the model. The same set of model constants was used for all cases and, through successive iterations, adequate agreement of computed and measured results was obtained. For this specific engine design and set of operating conditions, it was found necessary to account for the turbulence generated by the prechamber jet in order to reproduce main chamber combustion properly. In spite of uncertainties about both the measured quantities and the model, this first detailed comparison of engine combustion data also shows again that the model can be used to interpret and extrapolate engine data. Such capabilities are useful in the stages of development and optimization of engines.