We studied the mechanism of erythropoietin (EPO)-induced hypertension (HTN) in rats with chronic renal failure (CRF). After partial nephrectomy, rats were randomized into four groups. Group A received EPO, 150 U/kg, two times weekly for 6 wk to prevent anemia; group B received placebo injections and became anemic; group C received EPO but was kept anemic by dietary iron deficiency; and group D received placebo and regular transfusions to match hematocrit (Hct) in group A. Blood pressure (BP), Hct, platelet cytosolic calcium ([Ca2+]i) and magnesium concentration, and pressor and vasodilatory responses were determined. By design, Hct in groups A and D were comparable and significantly greater (P < 0.01) than in groups B and C. Despite divergent Hct values, the EPO-treated groups A and C showed a significant rise in BP compared with the placebo-treated groups B and D. HTN occurred whether EPO therapy was begun immediately or 4 wk after nephrectomy. EPO therapy augmented the elevation of basal [Ca2+]i and restored the defective thrombin-mediated rise of platelet [Ca2+]i in CRF animals. EPO therapy did not alter caudal artery contraction in response to either 68 mM K(+)-induced depolarization, angiotensin II or alpha 1-agonist, methoxamine in vitro, or the pressor response to angiotensin II in vivo. However, EPO therapy impaired the hypotensive response to nitric oxide (NO) donors, sodium nitroprusside and S-nitroso-N-acetyl-D,L-penicillamine, and reversed the CRF-induced upregulation of guanosine 3',5'-cyclic monophosphate production by thoracic aorta in vitro. Thus EPO-induced HTN in CRF rats is Hct independent and is associated with and perhaps causally related to increased basal and stimulated [Ca2+]i and impaired vasodilatory response to NO.