Fifty-four per cent of hospital stay days on the medical wards in a large teaching hospital, 40% on surgery and 26% on pediatrics, were found to be unjustified, using an assessment method based on explicit criteria. Despite the observed differences between the various specialties, the frequency distribution of the reasons for unjustified hospital stay days was essentially similar. Most of the unjustified hospital stay days were attributed to diagnostic evaluation, non-emergency treatment and immediate follow-up, which were carried out within the hospital rather than on an ambulatory basis. The major factors for unjustified hospital stay days were age, ethnic group, ward of admission and diagnostic category. These findings may contribute to the implementation of hospital utilization management programs aimed at cost containment.