PURPOSE: To survey graduates in practice from the first four classes of the University of New Mexico School of Medicine's (UNMSOM's) parallel curricular tracks, and compare data about the graduates' practice patterns, learning behaviors, and satisfaction with the profession of medicine. METHOD: Between 1979 and 1993, the UNMSOM had two tracks for the first two years of medical school: a conventional track and the Primary Care Curriculum (PCC), a community-oriented, problem-based track. In 1990, a survey was conducted of the 140 graduates from the first four classes (1983-1986) who had completed their postgraduate training: 40 from the PCC and 100 from the conventional track. Statistical methods included two-way analyses of variance, logistic regression, and chi-square, adjusted by Bonferroni methods. Comparisons between tracks are reported after adjustments were made for specialty effects. RESULTS: Thirty-three graduates (83%) from the PCC and 87 (87%) from the conventional tracks responded. The PCC graduates were much more likely to work in medically underserved areas, practice in publicly funded health care settings, and care for non-paying patients. The PCC graduates more often identified patient problems and curiosity as providing motivation for their learning. They more frequently studied clinical medicine and community health topics and spent time in community activities. The PCC graduates felt better prepared for practice by their undergraduate medical education. There was no difference between the graduates of the two tracks in the sizes of the populations in which they practiced, in the criteria they used for deciding on referrals to other physicians, in the ranges of community resource utilization, or in the degrees of satisfaction within their chosen professions. Large percentages of graduates from both tracks (67% conventional and 79% PCC) considered themselves to be practicing either primary care or a combination of primary care and non-primary care. In addition, 38% of all the graduates practiced in the state of New Mexico. More PCC graduates chose careers in family practice; however, no significant difference was found in a comparison between the proportions of PCC and conventional-track graduates who chose primary care careers. CONCLUSION: Track differences favorable to the PCC were evident in relation to the two major goals established by the program: to attract graduates to careers in primary care in rural and underserved areas and to provide graduates with self-directed, lifelong learning skills. Some expected track effects were not found.