Objectives: To measure the frequency of genotypes of the HFE ( haemochromatosis) gene in patients recruited to the west of Scotland coronary prevention study (WOSCOPS), and relate them to the subsequent occurrence of coronary clinical events. Design: Nested case - control study, drawing samples of DNA from the biological bank of a cohort study. Patients: Men aged 45 - 64 years in 1989, with moderate hypercholesterolaemia and no evidence of coronary heart disease at baseline. Interventions: Follow up for a mean period of 4.9 years. Typing for C282Y and H63D mutations of the HFE gene in 482 subjects with a subsequent coronary event and 1104 without an event. Results: The C282Y mutation was present in 81 of 482 cases (16.8%) and 182 of 1104 controls (16.5%). Comparing the prevalence of gene mutations in the cases and controls, there were no significant differences. The hazard ratio for C282Y heterozygotes was 1.03 (95% confidence interval (CI) 0.77 to 1.36) and for C282Y/H63D compound heterozygotes 1.04 ( 95% CI 0.50 to 2.14). Prespecified subgroup analyses of the pravastatin, placebo, smoking, and non-smoking groups showed no significant differences between cases and controls. Repeating the analyses after adjusting for possible confounding factors produced no change in the results. Conclusions: In a population of moderately hypercholesterolaemic middle aged Scottish men who did not have any evidence of coronary heart disease at baseline, the presence of a C282Y mutation in the HFE gene did not predict the occurrence of coronary events over a mean follow up of 4.9 years.