This paper describes a distinct approach in making technological forecasts. From historical data, fuel economy learning curves have been constructed for alternatives such as the Diesel, gas turbine, Stirling, stratified charge, and rotary engines. Assuming that evolutionary development will take place, projections of these learning curves are extended to the 1990 era. The investigation found that no engine is likely to exceed the fuel economy of the Diesel in the next 10-15 years. However, serious questions are being raised regarding the feasibility of controlling nitrogen oxides and unregulated emissions in future Diesel engines.