Newble et al. (1982) examined the construct validity of one Patient Management Problem (PMP) with groups of subjects with varied levels of medical competence, ranging from fourth-year medical students to post-intern medical registrars and consultant physicians. The present study reanalysed their data using analysis of variance followed by Scheffé a postiori comparisons of the mean scores for each group. Results suggest virtually no difference among the various groups' performances on this particular PMP, in contrast to the assertion of Newble et al. (1982) that 'the most competent groups (the post-interns) generally scored less well on the calculated indices than the senior students and interns'. While Newble et al. (1982) maintained that their findings do not show the PMP to be a valid test of competence, the present analysis suggests that their study was not an adequate test of the construct validity of PMPs due to lack of power and insufficient sample size. An alternative interpretation of this data is presented, along with implications for medical education and future research.