Sole fed commercial pellets suffers from a nutritional anaemia. The hypotheses tested are: (1) the nutritional anaemia in sole fed commercial pellets is caused by an iron deficiency; (2) the assumed iron deficiency is due to inadequate absorption of iron; (3) an increase in absorption due to a higher bioavailability of heme or iron chelates will alleviate anaemia in sole and (4) haematocrit (Hct) and haemoglobin (Hb) are expected to follow the iron absorption patterns. In addition, we estimated the absorption of copper, cobalt, chromium, manganese, molybdenum and zinc in order to evaluate possible interaction between available iron sources and the absorption of other divalent minerals. Sole were fed four diets, each with a different iron source: iron sulphate, iron methionate, iron proteinate or heme. Feeding was restricted and equal for all diets. At the start of the experiment, sole reared on commercial pellets had average values of Hct, Hb and the hepatosomatic index of 12.5%, 19.6 g.l- 1 and 1.13%, respectively. The values at the end of the experiment did not differ from the values at the start and were not affected by the source of iron. The apparent absorption coefficients (AAC) of iron, manganese, zinc, cobalt, chromium and molybdenum, except for copper, were unaffected by the iron sources. Yet, the iron absorption was high for all sources tested. The AAC of copper was 15–22% higher in sole fed the diet with heme. The use of the different iron sources, including heme, did not affect Hct and Hb in anaemic sole. The high absorption of iron and copper in sole fed heme did not affect Hct and Hb, which suggests that the nutritional anaemia in sole is not an iron nor a copper deficiency anaemia.