This study uses New York State hospital discharge data to examine the relationship between in-hospital mortality for a patient receiving an abdominal aortic aneurysm resection and the volume of aneurysm operations performed in the previous year at the hospital where the operation took place and by the surgeon performing the operation. Previous research on this topic is extended in several respects: (1) A three-year data base is used to examine the manner in which hospital and surgeon volume jointly affect mortality rate and to examine ruptured and unruptured aneurysms separately; (2) a six-year data base is used to study the "practice makes perfect" hypothesis and the "selective referral" hypothesis; and (3) the degree of specialization of high-volume surgeons is contrasted with that of other surgeons. The results demonstrate a significant inverse relationship between hospital volume and mortality rate for unruptured aneurysms. Further, very few surgeons substantially increased their aneurysm surgery volumes in the six-year study period. Weak selective referral effects were found for both surgeons and hospitals, and higher-volume aneurysm surgeons tended to have much higher specialization rates.