Studies that are the basis of recommended volume thresholds for CABG surgery are outdated and not reflective of recent advances in the field. This study examines both hospital and surgeon volume-mortality relations for CABG surgery through the use of a population-based clinical data set. Data from New York's clinical CABG surgery registry from 1997 to 1999 (total number of procedures, 57 150) were used to examine the individual and combined impact of annual hospital volume and annual surgeon volume on in-hospital mortality rates after adjusting for differences in severity of illness. Significantly lower risk-adjusted mortality rates occurred above all annual hospital volume thresholds between 200 and 800 and above all surgeon volume thresholds between 50 and 200. The number needed to treat (NNT) at higher-volume providers to avoid a death was minimized for a hospital threshold volume of 100 (NNT=50) and a surgeon threshold volume of 50 (NNT=118). The risk-adjusted mortality rate (RAMR) for patients undergoing surgery performed by surgeons with volumes of > or =125 in hospitals with volumes of > or =600 was 1.89%. The RAMR was significantly higher (2.67%) for patients undergoing surgery performed by surgeons with volumes of <125 in hospitals with volumes of <600. Higher-volume surgeons and hospitals continue to have lower risk-adjusted mortality rates, and patients undergoing surgery performed by higher-volume surgeons in higher-volume hospitals have the lowest mortality rates.