In the study reported here, the authors measured the differences between teaching and nonteaching hospitals in case mix complexity and provided some detail concerning the differences. They measured case complexity in a sample of 200 short-term general hospitals by the Resource Need Index (RNI) using a cross-classification of 3,490 case types with weights compiled from patient charges. Median RNI values were moderately higher for teaching than for nonteaching hospitals both for the hospital as a whole and for each clinical service except obstetrics-gynecology. The most resource-intensive case type were relatively more frequent in the teaching hospitals, but the least resource-intensive types were of about equal relative frequency in the two hospital groups. The results show that teaching hospitals could be expected to cost somewhat more per patient even if case mix were the only factor.