There are data suggesting that recombinant human erythropoietin (rHuEPO) may induce thromboses in hemodialysis patients, possibly due to alterations in platelet function. In an earlier study, we found evidence of platelet hyperfunction in several patients 4 to 8 weeks following the start of rHuEPO therapy, which was begun shortly after hemodialysis was initiated. Studies were performed to examine the effects of rHuEPO on whole blood platelet aggregation and adenosine triphosphate (ATP) release independent of changes in hematocrit or the uremic state. Eight hemodialysis patients without and four with a history of vascular access clotting had platelet aggregation tests performed at baseline while receiving rHuEPO, off rHuEPO for 2 weeks, and 4 to 6 weeks after restarting the drug. While the plasma EPO level decreased significantly after the 2-week period off rHuEPO (P < 0.0001), the hematocrit did not change at any of the three time periods. Whole blood platelet aggregation in response to adenosine diphosphate (ADP), collagen, and ristocetin was not significantly altered on or off rHuEPO in either patient group. Platelet hyperfunction, determined by aggregation or ATP release either spontaneously or in response to low-dose ADP or ristocetin, was not seen in any patient. These data suggest that the increase in access clotting is not the result of platelet hyperfunction induced directly by rHuEPO.