The polycythemic repression of erythropoiesis and the restoration of erythropoiesis by specific stimulation were studied in the spleen colony system in irradiated mice. 1. A 5 day period of erythropoietin stimulation (exogenous erythropoietin) or repression (polycythemia) of the bone marrow donor only, does not significantly alter the number or type of colonies formed by the transplanted marrow cells. 2. Erythropoietin stimulation did not alter the number or type of endogenous colonies formed in mice receiving 580 R. Erythropoietin repression (polycythemia) markedly reduced the growth but not the number of erythroid colonies, while not affecting the other types of colonies formed endogenously. 3. Erythropoietin stimulation of the irradiated recipient during colony growth did not alter the number or type of spleen colonies formed by transplanted marrow. Erythropoietin repression by polycythemia during colony growth completely suppressed the appearance of morphologically erythroid colonies without significantly altering the incidence of the other colony types. This effect of polycythemia was completely prevented by exogenous erythropoietin. Irradiated mice are therefore presumed to be secreting sufficient erythropoietin for maximal erythroid colony development. 4. The erythroid colonies suppressed by polycythemia were recognizable as microscopic foci of undifferentiated cells. Exposure of these foci to erythropoietin stimulation at different periods in their development was manifested by different degrees of growth and differentiation, from which it is apparent that erythropoietin stimulates not only morphological differentiation but also rapid mitosis. Retransplantation of either erythroid or of neutrophilic primary spleen colonies gave rise to both erythroid and neutrophilic secondary spleen colonies. The percentage of erythroid secondary colonies was slightly but significantly higher among the progeny of transplanted erythroid primary colonies than among the progeny of transplanted neutrophilic primary colonies. On the basis of these and other results, a working hypothesis is proposed for factors controlling the growth and differentiation of spleen colonies from transplanted bone marrow. It is postulated that most but perhaps not all spleen colony-forming units are pluripotent hemopoietic stem cells. It is further postulated that hemopoietic-inductive microenvironments (HIM) of different kinds exist in both the spleen and the bone marrow, and that these determine the differentiation of pluripotent stem cells into each of the lines of hemopoietic differentiation. Erythropoietin therefore may "induce" erythroid differentiation of only those stem cells under the influence of an erythroid HIM. Alternatively erythropoietin may act only as a growth and function stimulant of those stem cells that have been "induced" by an erythroid HIM into a state of erythropoietin responsiveness. In the latter case morphological differentiation presumably results from the functional activity stimulated by ESF.