Appropriate selection of medical students is a fundamental prerequisite if medical schools are to produce competent and caring doctors. The selection criteria for entry to the medical degree course at the University of Newcastle, New South Wales, are unique in Australia. The purpose of this study was to identify admission criteria that may predict performance in the first postgraduate (intern) year. Performance ratings were obtained from the clinical supervisors of two graduating classes of Newcastle University medical students during their five terms in internship (first postgraduate year). At least one rating was obtained for 93% of interns. A subset analysis of interns with multiple ratings (57%) showed that combining previous study in both humanities and science before medical school entry was predictive of higher intern performance ratings. These interns were rated more favourably than those who had studied science alone. Moreover, students who had earlier studied both humanities and science were twice as likely to complete their medical degree as those who had studied science alone. Age, gender, admission interview results, written psychometric test scores, academic marks, and whether previous tertiary study had been undertaken prior to medical school entry were not predictive of intern performance ratings. Subject spread, including a background in humanities, is important for effective medical practice, at least in the immediate postgraduate period. Perhaps it is time to evaluate the admission criteria by which medical students are selected.