One of the goals of the Automotive Stirling Engine Development Program, sponsored by the Department of Energy and managed by NASA/Lewis Research Center, is to develop a rationale for predicting transient CVS cycle emissions from steady-state engine data. A technique is developed that integrates engine emissions as a function of fuel flow over a modelled CVS cycle to predict vehicle urban cycle results. Steady-state emissions data from three Mod I engines* burning unleaded gasoline are used to predict vehicle NOX, CO, and HC emissions. A total of 155 data points representing variations in engine power, excess air (λ), and the exhaust gas recirculation (EGR) schedule are utilized. Predictions are then compared to the results of nine urban CVS cycle tests of the Mod I/Lerma vehicle, and a conclusion is reached that very accurate predictions of vehicle NOX emissions are possible. CO and HC emissions are considerably higher than predicted due to extreme sensitivity of CO emissions to Lambda, the effect of heater head temperature, and failure of the engine to accurately reflect emissions during start-up.