Background: Educational theories predict conflicting results for the effect of increasing the authenticity of the teaching format of complex information on educational outcomes. We sought to determine the effect of increasingly authentic small-group, preclerkship teaching format on clerkship outcomes to further enlighten this debate. Summary: Students enrolled in a prospective randomized crossover trial that involved three content areas. For each content area, three teaching formats were tested. Participants were randomized to teaching format by content area. Clerkship outcomes were performance on an objective structured clinical exam, a DVD exam, internal medicine clerkship grades, and performance on the subject examination. The data were analyzed using a multivariate analysis of covariance. One hundred and thirty-three (78%) students participated. Teaching format did not have a statistically significant effect on any of the specified clerkship outcomes. However, number of patients seen was significantly associated with higher scores in respective outcomes by topic. Conclusions: Second-year teaching format did not directly influence subsequent clerkship performance. Our study adds to the literature by demonstrating that the authenticity of preclinical teaching format does not appear to matter for clerkship performance; however, the number of actual patients seen does appear to influence related clerkship outcomes.