This study considered whether occupational groups can be discriminated by unique patterns of antecedent characteristics of an academic and non-academic nature. Statistical materials about family background, educational status at matriculation, and in-college data were assembled and coded on quantitative scales for 3 groups of Yale graduates: (1) the Classes of 1931 and 1932, (2) the Classes of 1935 and 1937, and (3) the Class of 1959. The most recent occupational information available about each graduate was coded according to an occupational code. Of the 25 occupational categories large enough for analysis in terms of the antecedent data, 5 had statistically unique patterns of matriculation and in-college characteristics. These were: (1) chemists and physicists, (2) engineers, (3) college teachers, (4) lawyers, and (5) salesmen. Nine discriminant variables were identified. These included: (1) age, (2) father's occupation, (3) mother's occupation, (4) score on SAT, (5) senior year average grade, (6) scholastic honors received at Yale, (7) involvement in fraternities, (8) academic, sports and social clubs, and (9) dramatics, publications and musical organizations. There was too much overlap, however, to predict later membership in a vocational group. (AF)