16 202 men, aged 40-49 years, were screened for coronary risk factors, Of these, 1232 healthy, normotensive men at high risk of coronary heart disease (CHD) were selected for a 5-year randomised trial to show whether lowering of serum lipids and cessation of smoking could reduce the incidence of CHD. Men were admitted to the trial if they had serum cholesterol levels of 7.5-9.8 mmol/l (290-380) mg/dl), coronary risk scores (based on cholesterol levels, smoking habits, and blood pressure) in the upper quartile of the distribution, and systolic blood pressures below 150 mm Hg (mean of two measurements). The men in the intervention group were recommended to lower their blood lipids by change of diet and to stop smoking. Mean serum cholesterol concentrations were approximately 13% lower in the intervention group than in the control group during the trial (based on the difference between the mean of 3 prerandomisation values and the mean of yearly values during the trial). Mean fasting serum triglycerides fell by 20% in the intervention group compared with controls. 80% of the men in both groups smoked tobacco daily at the start of the study. The mean tobacco consumption per man decreased by 45% more in the intervention group than in the control group. However, only 25% of the smokers in the intervention group completely stopped smoking compared with 17% in the control group. Diagnosis of events of cardiovascular disease during the study was made blindly according to predefined criteria by two cardiologists not involved in the study. At the end of the observation period the incidence of myocardial infarction (fatal and non-fatal) and sudden death was 47% lower in the intervention group than in the controls (p = 0.028, two-tailed log rank test). When the incidence of strokes was added, the difference between the groups was still significant. It is concluded that in healthy middle-aged men at high risk of CHD advice to change eating habits and to stop smoking significantly reduced the incidence of the first event of myocardial infarction and sudden death. Statistical analysis, by Cox's proportional hazards model shows that the reduction in incidence in the intervention group is correlated with the reduction in total cholesterol and to a lesser extent with smoking reduction.