Considerable experimental and clinical evidence suggests a relationship between erythropoiesis and thrombopoiesis. This is supported by observations that erythropoietin (Epo), the primary regulator of erythropoiesis, can affect platelet production when injected into animals. In this study we provide experimental evidence for a direct effect of Epo on thrombopoiesis by demonstrating that 125I-labeled recombinant human Epo binds to rat and mouse bone marrow megakaryocytes. Thus, autoradiographic analysis using cold competition to measure specific binding has been used to demonstrate that Epo binding to megakaryocytes increases with megakaryocyte maturation. When corrected for cell size, Epo binding sites per unit surface area increase from Stage I megakaryoblasts to Stage II megakaryocytes, and then remain approximately constant throughout further megakaryocyte maturation. Receptor density on megakaryocytes is similar to that on pronormoblasts in the rat, and in mice is 60% that on pronormoblasts. No binding of Epo to platelets or to naked megakaryocyte nuclei was detected. Equilibrium binding studies with partially purified rat megakaryocytes (20%-40% pure), where megakaryocytes are the only significant Epo binding cell population, showed a single class of saturable, high-affinity binding sites present on average at 6500 binding sites per megakaryocyte with a KD of 287 pM. Binding of [125I]Epo to rat megakaryocytes was inhibited with an antiserum against murine erythroblasts. These results suggest that the effects of Epo on thrombopoiesis may be directly mediated through specific, high-affinity binding sites for Epo on the surface of maturing megakaryocytes.