Changes in the composition of medical school classes, in methods of personality assessment, and in the nature of medical specialties motivated the authors to reexamine the relationship between a student's personality and his or her chosen medical specialty. In the mid-1980s, the authors assessed two consecutive classes of medical school students (120 in each class) on 15 personality measures throughout their four years; usable data were obtained for 199 of the students. The students' specialty choices upon graduation were obtained from the results of the National Resident Matching Program, and were classified as follows: internal medicine, surgery, family practice, pediatrics, psychiatry, obstetrics-gynecology, and hospital-based specialties. The authors constructed personality profiles for the students who had chosen each of these seven specialty groupings, on the basis of t-tests and discriminant function analyses, and discuss these profiles and their associations with the students' specialty choices in the context of earlier research. The personality profiles for surgeons were remarkable for their consistency with reports from the literature from the 1950s and 1960s, while the profiles for obstetricians, psychiatrists, and pediatricians were complex and provocative. Distinctive personality profiles did not emerge for internists or family practitioners.