The aim of this study was to examine whether observed high mortality from ischaemic heart disease (IHD) among males in the municipality of Hofors, Sweden, is accurate, or the result of some artefact. The validity of death certificates from Hofors municipality and from a control area (Gävle) were ascertained by comparing them with medical records from hospitals and district physician offices, and reports from clinical or medicolegal autopsies. The results indicate that the observed high mortality from IHD among Hofors males cannot be explained by deficiencies in establishing the cause of death. Results from other validity studies indicate, however, that there is a risk of local variation when reporting causes of death. In many instances this motivates checking the validity of data before they are incorporated into official statistics and used as a basis for planning.