OBJECTIVE We assessed the operative mortality of coronary artery bypass grafting (CABG) surgery using risk stratification. METHODS In 294 consecutive patients who underwent CABG with or without concomitant surgery from August 1994 to December 1999, we compared operative mortality calculated conventionally and by risk stratification. Scores for each patient were calculated using the Parsonnet additive model and stratified based on the probability of operative mortality. RESULTS Overall crude hospital mortality was 4.8%-4.0% among patients younger than 80 years and 14% among those 80 years of age or older (p = 0.0692). Hospital mortality was 12% in urgent/emergency surgery, and 1.5% in elective surgery (p < 0.0002), and 4.5% in CABG alone and 7.4% in CABG with concomitant surgery (p = 0.3763), and 25% in patients receiving vein grafts only and 3.0% in those receiving at least 1 artery graft (p = 0.0003). Overall patient distribution was 32% good, 20% fair, 20% poor, 11% high-risk, and 16% extremely high-risk. Predicted mortality was 2.2% for patients who were a good risk, 6.7% for fair-risk, 12% for poor-risk, 16% for high-risk, and 25% for extremely high-risk patients. Actual operative mortality was 1.0% for good-risk, 0% for fair-risk, 3.4% for poor-risk, 6.3% for high-risk, and 18% for extremely high-risk patients, making actual mortality significantly lower than that predicted. CONCLUSIONS Comparing predicted mortality and actual mortality enabled us to objectively calculate operative results and assess operative quality.