The purpose of this study was to determine if internal medicine residency type or location was associated with differences in performance of candidates as measured by the internal medicine certifying examination. Included in the study were all first-time taker candidates for the 1986 to 1996 American Osteopathic Board of Internal Medicine certifying examinations in internal medicine. Group analysis was performed based on the type of residency track leading to board eligibility: (1) traditional internship plus 3 years of internal medicine residency; (2) traditional internship plus 2 years of internal medicine residency and 1 year of subspecialty training; (3) specialty track internship plus 2 years of internal medicine residency; and (4) traditional internship plus 3 years of allopathic internal medicine residency. Subgroup analysis of the subspecialty track group was performed to determine if any particular subspecialty-trained subgroup performed better than the others. Results indicate that all groups had similar scores and pass rates except for the allopathic-trained residents, whose scores and pass rates were lower. Subgroup analysis of the subspecialty-trained candidates revealed that procedure-oriented subspecialty candidates performed similar to nonprocedure-oriented candidates. No gender differences were noted in scores or pass , rates. It is concluded that the overall performance of candidates is equivalent for . each of the residency training tracks developed by the American College of Osteopathic Internists. Future performance on the recertification examination will need to be tracked to determine if these trends continue for practicing internists.