PURPOSE: To elucidate the link between the quantity and quality of clinical exposure gained by first-year clinical students in hospital settings and their performance on a subsequent comprehensive assessment of clinical skills (the objective structured clinical examination, or OSCE). METHOD: Data relating to educational activities and workload were collected for the second introductory clinical attachment undertaken by 152 (of 246) students in two British medical colleges prior to a joint comprehensive 22-station OSCE administered in May 1994. Pearson correlation coefficients were used as the main analytical tool to study the relationships between measures of clinical activity and total OSCE scores. RESULTS: In general, of 43 indices of the amount, nature, and quality of bedside, ward-based, or outpatient experience, only six correlated with OSCE scores. The strongest links were for whether students examined out-patients on their own (r = .2), whether the objectives had been made clear (r = .19) and the number of clinics attended (r = .18). Variables meeting the criteria were entered into a backwards stepwise regression analysis to predict total OSCE scores, but they explained only 23% of the variance. CONCLUSION: The association between clinical experience and educational outcomes remains poorly understood.