The EPA certification data base is the most extensive available body of information relating engineering variables to fuel economy. The range and distributions of these variables are presented in a non-sales weighted context. A multiple regression analysis, performed with careful attention to selection of optimum non-linear forms for some of the variables, shows that the most significant determinant of fuel economy is the parameter CID × N/V. Vehicle weight is second.A sensitivity study on the isolated fuel economy effects of CID, N/V and weight confirms the superiority of CID as an influencing factor, and opens avenues for improving the structure of the forms of these variables in future regression equations.A method is devised for quantifying ignition timing over a driving cycle in terms of a single “labeling” value usable as an engineering variable. Although not illustrated, the method is applicable to carburetion in similar fashion.Relative fuel economy comparisons drawn between Federal (49-States) and California certification cars from the 1974, 1975 and 1976 model years show that an apparent inverse relationship between fuel economy and the stringency of emission controls can be overcome with the concurrent passage of time and the evolutionary application of improved technology.