BACKGROUND: Male gender is an independent coronary risk factor. METHOD: Long-term follow-up of 989 Danish men who underwent legal castration between 1929 and 1968. RESULTS: The legally castrated men were unmarried and belonged to social class IV and V more often than were Danish men in general. During the follow-up until 2000, 835 of the 989 (85%) castrated men died, including 148 who died of myocardial infarction. In multiple Poisson regression analyses, the men had a standardized mortality rate (SMR) for all-cause mortality of 1.30 (95% CI: 1.26-1.36) and a SMR for mortality of myocardial infarction of 1.08 (95% CI: 1.04-1.16). Thus, the castrated men had a lower proportion of deaths of myocardial infarction (148/792, 18.7% (95% CI: 16.0-21.6%)) than was expected based on the mortality rates for the Danish male population (136/608, 22.4%). The castrated men had discordant changes for the SMR for all-cause mortality and mortality of myocardial infarction whereas subgroups of the Danish population previously has been found to have concordant changes for the two SMRs. CONCLUSION: The castrated men had fewer deaths of myocardial infarction than expected, so men may not have increased risk of coronary heart disease from unphysiologically low levels of endogenous androgens.