The duration of the reduction of mortality after coronary artery bypass grafting (CABG) is an important issue and this study was undertaken to evaluate time in relation to excess mortality among CABG patients compared with the general population. Survival was analysed in 4661 patients who had undergone their first isolated CABG. Observed survival was related to that expected among subjects from the general Swedish population stratified by age, sex and 5-year calendar period, to calculate relative survival and estimate disease-specific survival. Relative survival (including all deaths) was 94.6% at 5 years, 82.5% at 10 years, and 59.9% at 15 years. A multivariate model based on relative survival rates adjusted for age, year of surgery, severity of coronary disease, left ventricular function, and smoking habits was used. Compared with the first year of follow-up, the relative hazard (a measure of the risk of death) was at a minimum 2 years after surgery, but was dramatically increased after about 8 years. Relative survival was worsened by smoking at the time of operation and by moderate or severe left ventricular dysfunction pre-operatively. The survival rate was higher among patients operated on after 1985 than among those operated on earlier.