The 12 acute care public hospitals in New York City (HHC hospitals) are compared with a matched group of nonpublic hospitals (non-HHC hospitals). The following questions are considered: using DRGs to define case mix, how does the case mix of HHC and non-HHC hospitals differ; to what extent do differences in case mix account for differences in average length of stay (ALOS); can factors other than case mix be identified that may explain differences in ALOS? Although about one half of the 20 most prevalent DRGs are similar in both types of institutions, there are clear case mix differences. The higher percentage of abortion, psychiatric, and chemical dependence discharges and the lower prevalence of surgery in HHC hospitals contribute to this difference. The case load is more concentrated in fewer DRGs in HHC hospitals. HHC hospitals treat more patients in DRGs with a shorter ALOS, but, on the average, patients in the same DRG stay more than 1 day longer in HHC facilities. To some extent, this longer LOS can be explained by differences in payor type, primary diagnosis within a DRG, and, most important, in the percentage of outlier patients.