Therapeutic venesection for polycythaemia vera (PV) produces iron deficiency. If iron supplementation is avoided, the frequency of venesections can be kept lower than if iron is given. This is the standard treatment for PV in our department, and this model for iron deficiency was used to compare serum ferritin, free erythrocyte protoporphyrin (FEP), serum iron and transferrin as indicators of iron deficiency. Eleven patients with PV were studied on a total of 90 occasions. Five patients were followed from normal iron status to iron deficiency, the other six were iron deficient at the start of the study. Serum ferritin and FEP became abnormal approximately simultaneously during the development of iron deficiency, serum ferritin in all patients, FEP in 8 out of 11 patients. There was a correlation between the two in all specimens (r = 0.75, p less than 0.001), but serum ferritin showed fewer false negative results. Serum transferrin alone was elevated only in 25% of the patients, and serum iron, although mostly subnormal, was rather inconsistent. It is concluded that serum ferritin and FEP can both be used to diagnose iron deficiency during venesection treatment of PV, whereas serum iron and transferrin are of little value.