Background: To investigate the predictive validity of the Medical College Admission Test (MCAT) for clinical reasoning skills upon completion of medical school. Method: A total of 597 students (295 males, 49.4%; 302 females, 50.6%) participated from 1991 to 1999. Stepwise multiple regressions of the MCAT and premedical school GPA (independent variables) on the Part 1(declarative knowledge) and Part 2 (clinical reasoning) of the Medical Council of Canada Examinations (dependent variables) were employed. Results: For Part 1, the multiple regression revealed that three predictors (verbal reasoning, biological sciences, GPA) accounted for 23.3% of the variance, and for Part 2, two predictors (verbal reasoning, GPA) accounted for 11.2%. Conclusion: There is both convergent and divergent evidence for the predictive validity of the MCAT for clinical reasoning.