Three hundred and one clinical medical students in four universities took the same 50 question MCQ dermatology examination after their dermatology teaching. In one centre, half the students had had additional teaching; these students performed better (mean score 47.5%, n = 29) than those who had no extra teaching (mean score 40.9%, n = 29). In another centre, the students' mean score improved from 24.1 (SD = 6.7) before to 41.6 (SD = 7) (n = 46, P < 0.001) after their dermatology teaching. The different subject areas covered by the examination were analysed separately. In the lowest scoring centre (mean score 34.0, SD = 9.4) the students scored lowest in 9 of the 14 subject areas. In the highest scoring centre (mean score 47.5, SD = 9.9) students scored highest in 7 of these 14 subject areas. This study enabled questions of high discriminatory value to be identified for future use. The use of the same examination in different centres provides feedback for the centres concerning strengths and weaknesses of their teaching.