The purpose of this study was to discover if predictive variables could be identified to assist manpower programs with decisions concerning trainee program assignment. The primary problem explored was: What is the relationship between selected trainee characteristics, specific vocational assessment procedures and successful outcomes in a CETA skills training center? The research concentrated on demographic variables of age, sex, ethnic group membership, education and family status, as well as scores on three sp~cific assessment procedures: academic tests, work samples, and a work behavior rating instrument. The study involved a sample of 137 subjects who had completed a vocational assessment experience, had enrolled in one of four vocational classes, and had been terminated, either successfully or unsuccessfully, from a skills training program at the Northern Virginia CETA Skill Center during fiscal year 1981-82. Demographic and test data were analyzed by means of a discriminate analysis program in an effort to differentiate between successful and unsuccessful trainee termination. The results of the study indicated that for each of the three assessment techniques tested: (1) Academic Tests, (2) Work Samples, and (3) Work Related Behavioral Observations, several significant variables were identified. Results of the discriminant analysis were astonishingly h:i.gh and appear to demonstrate that statistical relationships can be proven to exist between certain trainee characteristics, assessment procedures and training outcomes. When the discriminant analysis using the various demographic variables, scores on the work behavior rating instrument, academic tests and work samples were performed, the results were impressive. The discriminant analysis yielded predictive accuracies of 81%-100% for successful and unsuccessful terminatins in three training area. Thus it would appear that the specific demographic variables highlighted by the three assessment techniques employed at the Northern Virginia CETA Program are valuable indicators of trainee outcome. / Ed. D.