The multivariate statistical procedure has been used to compare various alternate fuels based on the long term engine performance. The results from the multivariate analysis of covariance procedure, which does the simultaneous comparison of the chosen response variables, were adjusted for time of engine operation since the time factor was found to be significant. The assumptions for this test procedure are presented. Problems which could affect the validity of the statistical procedures for the laboratory endurance test are discussed.To illustrate the statistical method, data from a laboratory screening endurance test was used. Results of the analysis were in close agreement with the engineering interpretation of the observed differences between fuels.