Erythroid burst-forming units (BFU-E) from mouse bone marrow were grown for 7 days in agar or serumfree methylcellulose cultures in the presence or absence of erythropoietin (Ep) and/or interleukin 3 (IL-3). It was found that IL-3, even in the absence of serum and detectable Ep, was able to stimulate the full development of many erythroid bursts. This IL-3 effect was cell-dose dependent and did not appear to correlate with Ep dose. Spontaneous bursts and those stimulated by Ep only were rare and when seen were very small relative to those produced by IL-3 or IL-3 plus Ep. When addition of IL-3 or Ep to 7-day cultures was delayed, IL-3 but not Ep was shown to maintain BFU-E. No evidence was found by radioimmunoassay that Ep was produced or released in 7-day, "serum-free" cultures of bone marrow nor was Ep activity detected in culture media except those to which it had been added deliberately.Interleukin 3 (IL-3) began its interesting history as a factor believed to regulate early T-cell differentiation. This was based on the observation by Ihle that a factor in conditioned medium from activated T cells had the ability to induce expression of 20a-hydroxysteroid-dehydrogenase, an enzyme uniquely associated with the T-cell lineage, in splenic lymphocytes from athymic nude mice (1, 2). Ihle et al. (3) proposed that this factor, a glycoprotein of Mr 28,000, be called IL-3 and has since purified it to homogeneity (4). Activated T cells are believed to be the major source of IL-3, although WEHI-3, classified as a myelomonocytic leukemic cell line, produces it constitutively. In the short time since WEHI-3 was shown to produce IL-3, it has been found that the factor also has the ability to stimulate the growth in vitro of many kinds of hemopoietic cells (5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19) and it has been called by some a multispecific colony-stimulating factor (20).It has been shown to be identical with Schrader's P-cellstimulating factor (21) and with stem cell-activating factor (22).It has been recognized for some time that early erythroid progenitors require a second factor, burst-promoting activity (BPA), in addition to the hormone erythropoietin (Ep) for their full development in culture (23-26). The ability of IL-3 to stimulate erythropoiesis has been attributed to its activity as a burst promoter in the production of differentiated erythrocytes in vitro from the burst-forming progenitor or erythroid burst-forming units (BFU-E) (17, 27). It has been reported that under some circumstances "Ep-independent" BFU-E can develop (27-34). Evaluation of the site and nature of action of both BPA and Ep have, until recently, been burdened by the lack of highly purified material and by the presence of unknown quantities of BPA and Ep in constituents of culture media. Some crude preparations of Ep have been found to contain BPA as a contaminant (35). Now that a completely "serum-free" system has been devised (36) for the in vitro growth of erythroid bursts, it is possible to seek experimental answ...