It has been suggested that teaching hospitals as a group have done well financially under Medicare's prospective payment system. If so, is this because teaching hospitals have reduced inefficiencies or because their case mix is not as severe as presumed? In this study, we used Disease Staging and diagnosis related groups (DRGs) to isolate case mix attributed to given patient populations from that attributed to hospital treatment standards. We also analyzed differences among types of teaching hospitals. We found few case-mix differences between teaching and nonteaching hospitals when the weighting system was independent of resource consumption (i.e., Disease Staging). However, when resources were used to weight case-mix measurement (i.e., DRGs), teaching hospitals were found to have a more serious case mix. We conclude that although teaching hospitals typically do not have a more severe case mix than nonteaching hospitals, they do use more resources to treat their patient mix under DRGs.