OBJECTIVE The objective structured clinical exam (OSCE) is increasingly being used to evaluate student clinical performance. However, scant literature exists pertinent to this approach in evaluating family medicine clerkship performance. In this study, we assess 8 years' experience with a family medicine clerkship OSCE. METHODS Eight annual clerkship OSCEs and the performance of 696 students are described. Comparisons of faculty evaluation, written exam, and OSCE performance are made for 335 students. Post-OSCE student and faculty feedback regarding OSCE validity and utility is also presented. RESULTS Student performance is highest in medical history taking and physical examination and lowest in information-sharing stations. OSCE results appear to be relatively consistent on a year-to-year basis. OSCE, faculty evaluation, and written exam results have low overall levels of correlation, particularly in assessing performance that differs substantially from the mean. Students and faculty agree that the OSCE experience reflects skills that students should possess, but there is less agreement that the OSCE reflects clerkship-related learning and actual student performance. Both students and faculty derive insight from the OSCE regarding the definition of specific learning needs. CONCLUSIONS The family medicine clerkship OSCE we describe appears to provide consistent measures of student performance. Although content validity is high, further assessment is needed to assure construct validity. The OSCE experience provides students with a rich resource for defining clerkship-related learning needs. Study results strongly suggest that OSCEs, faculty evaluations, and written exams provide differing measures of student performance. The reasons for these differences merit further exploration.