Although there is less attrition from medical school now than in prior years, 500 or more students can be expected to withdraw from American medical schools during each academic year. Continuing study of the problem is, therefore, warranted. Personality inventory, cognitive, and application data, all gathered at the time of admission, were examined for 1,014 graduates and 57 dropouts; separate analyses were conducted for the 17 who departed for academic reasons and the 40 whose attrition was attributable to other causes. Correlational analyses were also conducted in which graduates, nonacademic dropouts, and academic dropouts were scaled in a 4-3-1 continuum. The best predictor of this graduation versus dropout hierarchy was given by a six-variable combination, including scores on the Quantitative ability subtest of the Medical College Admission Test and premedical grades for the last two terms, with positive weightings; personality inventory scales for status potential, socialization, and communality, also with positive weightings; and a personality scale for conformist achievement drive, weighted negatively.