The role of an elevated serum leukocyte count (WBC) as a coronary risk factor was investigated using a nested case-control design in dyslipidemic middle-aged men (n = 420) participating in the Helsinki Heart Study, a coronary primary prevention trial. Baseline WBC was significantly higher, 6.93 (2.11) x 10(9)/L in subjects with cardiac events, than in controls, 6.26 (1.88) x 10(9)/L; p less than 0.002. This association was time-dependent, however, since the difference was not significant for events occurring during the second half of the 5-year study. Using nonsmokers in the lowest WBC tertile as the reference sample, the relative risks in the highest WBC tertile were 1.86 (95% confidence intervals [CI] 0.81 to 4.28) for nonsmokers and 3.07 (95% CI 2.23 to 8.19) for smokers. Logistic regression analysis including smoking in the model disclosed an independent contribution of elevated WBC to coronary heart disease. We conclude that elevated leukocyte count was a coronary risk factor even in this dyslipidemic population.