Studies with anaemic children and pregnant women from areas where vitamin A deficiency is endemic have shown a beneficial effect on Fe status of supplemental vitamin A in addition to Fe supplementation. This suggests a relationship between vitamin A and Fe status, which we attempted to mimic in rats with anaemia and chronic vitamin A deficiency. Male rats were fed on Fe-adequate diets (35 mg Fe/kg) containing different levels of vitamin A (1200, 450, 150, 75 and 0 retinol equivalent (RE)/kg feed) until they were 5 weeks old. These diets were identical to the diets fed to their mothers. Then the young male rats were transferred to diets containing the same levels of vitamin A but no added Fe. After another 2 weeks the rats were repleted with Fe (35 mg/kg feed) without or with vitamin A to a level of 1200 RE/kg feed. Increased vitamin A intake by the groups previously fed on diets with either 0 or 75 RE/kg produced a reduction in blood haemoglobin concentration, packed cell volume and erythrocyte count. In the group which had been fed on the diet without vitamin A, supplemental vitamin A raised mean cell volume, plasma Fe concentration and total Fe-binding capacity. Vitamin A supplementation during the period of Fe repletion produced a decrease in splenic and tibia Fe concentration, the effect being greater with increasing severity of previous vitamin A deficiency. The paradoxical effect of supplemental vitamin A on haemoglobin, packed cell volume and erythrocyte count can be explained by a decrease in the degree of haemoconcentration. Thus, the positive effect of supplemental vitamin A seen in humans is also observed with rats under controlled experimental conditions. We speculate that supplemental vitamin A during Fe repletion contributes to optimum erythropoiesis and Fe mobilization when baseline vitamin A status is impaired.