Weanling male Wistar rats were fed a low iron basal diet for 4 weeks. They were then divided into 16 groups according to a randomized block design based on hemoglobin level. During the repletion period of 2 weeks, one group was fed the basal diet. The other fifteen groups received ferrous sulphate, freeze-dried ground beef, ground granular soy product, a 2:1 mixture of beef and the soy product, or rapeseed protein concentrate as iron sources at three levels. All diets were isocaloric and isonitrogenous. The efficiency of iron utilization was calculated on the basis of iron gained in hemoglobin and iron consumed, applying appropriate correction for iron obtained from the ingredients of the basal diet. The relative biological value (RBV) of iron in the test source was calculated as the ratio between the efficiency for the test and the standard source (ferrous sulphate). The RBVs were: freeze-dried ground beef, 0.53; ground soy product, 0.62; 2:1 mixture of beef and soy product, 0.71; rapeseed protein concentrate, 0.55. The low availability of iron in beef containing 60% heme iron indicated that the anemic rat is not a suitable model for normal man, since the absorption of heme iron by man is much higher than that of inorganic or non-heme iron.