A study was undertaken to investigate the validity of a progress test, the Maastricht Progress Test, that was designed to measure knowledge and clinical reasoning growth in a problem-based medical curriculum. Scores and subscores of about 40 students per year (total sample of 195) on the different categories of the progress test were compared with scores on a clinical reasoning test. Both tests revealed the same pattern of increasing scores over the years, and they had a high correlation. Further analyses revealed that the clinical science component, in particular, of the progress test explained the variations in the clinical reasoning test scores. Knowledge of behavioral sciences had a small but independent contribution. Outcomes are discussed from the perspectives of research and theory on the development of medical expertise, and educational consequences are discussed. An appendix contains an example problem. (Contains one table, three figures, and four references.) (Author/SLD)