The iron hypothesis as an alternative explanation for the gender difference in the incidence and mortality of atherosclerosis has provoked increased debates and public health concerns. In this review we summarize the historical and recent literature on the iron hypothesis and discuss several related clinical issues and their implications. Apart from misconstruction of study populations, lack of a good method to reflect the iron contents of tissues may be the major factor for causing inconsistent results from epidemiological studies. Published data from 11 countries clearly indicate that the mortality from cardiovascular diseases is correlated with liver iron. We propose that redox-active iron in tissue is the atherogenic portion of total iron stores. Recently developed magnetic resonance imaging techniques in combination with Fe chelators may allow future studies to examine this component of body iron in lesions and the whole body. Several clinical situations characterized by increased iron stores have been proposed as 'human models' suitable for further tests of the iron hypothesis. Patients with end-stage renal disease may be the most unique cohort, having significant increases in their iron stores, low-density lipoprotein (LDL) oxidation, and cardiovascular events. Other patient groups may be well suited for specific studies of different atherogenic events. With a better understanding of iron-driven oxidative damage, well controlled and effectively designed studies on these models will finally bring us to the truth of the iron hypothesis.