Eleven-year mortality rates were studied in middle aged men who had participated in a randomised 5-year multifactorial primary prevention trial on cardiovascular diseases during 1974-1980. The men were given health education advice before the study. The 5-year trial markedly improved the risk factor status in the men in the intervention group (n = 612), but their 5-year incidence of total coronary events tended to be higher than in the randomised non-treated control group (n = 610) and significantly higher than in an non-randomised, non-treated low risk group (n = 593). During the six years following the discontinuation of the trial, 11 deaths from cardiovascular disease occurred both in the intervention and in the control groups and three in the non-randomised low risk group. Thus, the cumulative eleven-year cardiovascular mortality rates and their 95% confidence intervals (Cl95) were 2.45% (Cl95: 1.38, 3.67) in the intervention group and 1.97% (Cl95: 1.01, 3.34) in the randomised high risk control group. In the non-randomised low risk group the mortality rate was 0.51 (Cl95: 0.01, 1.46). Multiple logistic regression analysis showed that overweight and hypercholesterolaemia, and smoking in the high risk controls, were the initial risk factors associated with the 11-year cardiovascular mortality. The latter was not accumulated in any treatment measure during the prevention period. Furthermore, despite the unfavourable effect of beta-blocking agents on total cardiac events during the intervention, beta-blockers were not associated with cardiac deaths in the 11-year follow up.(ABSTRACT TRUNCATED AT 250 WORDS)