Theiron hypothesis'[1±6] suggests that iron depletion protects against ischemic heart disease, thus explaining several phenomena including the low risk observed in menstruating women. The hypothesis implies that protection is lost upon acquisition of stored iron following the cessation of regular menstrual bleeding. If the loss of iron depletion is related to development of ischemic heart disease, it is reasonable to ask how a genetic predisposition to increased iron storage influences the risk of cardiovascular disease. In the early years following publication of the iron hypothesis, it was often remarked that the apparent lack of increased ischemic heart disease in hereditary haemochromatosis made the hypothesis untenable. More recently, with the publication of studies showing increased cardiovascular event rates in haemochromatosis heterozygotes [7±9], it is clear that the hypothesis cannot be discarded on the basis of this argument. Additional studies [10±12] on atherosclerosis and cardiovascular event rates in heterozygotes present conflicting findings on the relationship between iron loading HFE genes and cardiovascular diseases. Franco et al.[10] and Rossi et al.[11] see no association of heterozygosity with measures of atherosclerosis, although Rossi et al.[11] report a positive association of carotid plaque with serum ferritin in women. In this issue of the European Journal of Clinical Investigation, a new study reports a strong positive association of heterozygosity with atherosclerosis in H63D carriers who smoke in one population, but negative associations with event rates in two other ongoing studies [12]. Discrepancies with previous findings of positive associations [7, 8] may be explainable in part on the basis of differences in study design [12]. However, there are also inherent difficulties in the epidemiological approach to the question of the association of cardiovascular diseases with heterozygosity [9], arising in particular from the almost complete overlap in the distributions of iron storage levels among heterozygotes and normal controls. We will first discuss three considerations in evaluating the relevance of homozygous haemochromatosis to the iron hypothesis, and then present a perspective on the meaning of the recent studies on heterozygosity and cardiovascular diseases.