There is still debate in the literature on whether or not endurance athletes tend to have low iron stores. In this article, we propose that endurance athletes really are at risk of becoming iron deficient due to an imbalance between absorption of dietary iron and exercise-induced iron loss. The purpose of this article is to present a critical review of the literature on iron supplementation in sport. The effect of iron deficiency on performance, its diagnosis and suggestions for treatment are also discussed. Studies of the nutritional status of athletes in various disciplines have shown that male, but not female, athletes clearly achieve the recommended dietary intake of iron (10 to 15 mg/day). This reflects the situation in the general population, with menstruating women being the main risk group for mild iron deficiency, even in developed countries. Whereas the benefit of iron supplementation in athletes with iron deficiency anaemia is well established, this is apparently not true for non-anaemic athletes who have exhausted iron stores alone (prelatent iron deficiency); most of the studies in the literature show no significant changes due to supplementation in the physical capacity of athletes with prelatent iron deficiency. However, the treatment protocols used in some of these studies do not meet the general recommendations for the optimal clinical management of iron deficiency, that is, with respect to adequate daily dosage, mode of administration and treatment period. For future studies, we recommend a prolonged treatment period (> or = 3 months) with standardised conditions of administration (use of a pharmaceutical iron preparation with known high bioavailability and a dosage of ferrous (Fe++) iron 100 mg/day, taken on an empty stomach). Currently, decisions regarding iron supplementation are best made on the basis of taking care of individual athletes. We believe that there are sufficient arguments to support controlled iron supplementation in all athletes with low serum ferritin levels. Firstly, the development of iron deficiency is prevented. Secondly, the nonspecific upregulation of intestinal metal ion absorption is reverted to normal, thus limiting the hyperabsorption of potentially toxic lead and cadmium even in individuals with mild iron deficiency.