Anemia of endstage renal failure improves shortly after kidney transplantation. However, in about 10% of transplanted patients polycythemia occurs. By use of a sensitive in vitro bioassay the pathogenetic role of erythropoietin (Ep) was investigated in 12 patients with post-transplant erythrocytosis (PTE), and compared to 12 non-PTE patients. The mean Ep of 160 mU/ml was significantly elevated in patients with PTE as compared to 25 mU/ml of 36 healthy controls, whereas, the mean Ep of 24 mU/ml in non-PTE patients did not differ significantly from healthy controls. To further elucidate the mechanism of inappropriate Ep production, selective venous catheterization of native and transplanted kidneys was performed in six patients. In four PTE patients the mean Ep in native kidney veins of 110 mU/ml was significantly higher than the peripheral Ep of 66 mU/ml, whereas, mean Ep in kidney graft veins was 51 mU/ml. In contrast, in two non-PTE patients no significant difference between mean Ep from native and transplanted kidney veins was observed. We conclude that some patients escape from normal feedback regulation either due to autonomous Ep production or due to feedback regulation at an elevated level of hematocrit and that inappropriate Ep production originates from the diseased native kidneys.