The general practitioner often requires a simple and reliable method of determining the potential risks of surgical intervention. We derived and tested a simple clinical scoring system for the preoperative prediction of 30-day mortality after coronary artery bypass surgery. From a national register of all open heart operations in the Republic of Ireland 1984-1989, we identified 4276 male patients who had primary isolated non-emergency coronary artery bypass surgery. Using logistic regression, we derived a clinical scoring system to predict operative (30-day) mortality in patients operated on between 1984 and 1987. We then prospectively evaluated the score on patients seen over the next two years. Variables identified for our scoring system were age, recent myocardial infarction, left ventricular failure, extensive distal coronary artery disease and body surface area. Five risk categories were defined; mortality in the high-risk group was 9.7-fold (95% CI: 4.6-20.7) greater than in the low-risk group. When tested on new patients, the relative mortality between the two risk groups was 15.2 (4.6-50.5). The observed and predicted mortalities in each risk group showed close agreement. This clinical scoring system, easily used by a general practitioner, can predict operative mortality in males for whom primary isolated coronary artery bypass surgery is contemplated.