We have recently discovered that cells of Coon's Buffalo rat liver (BRL) line secrete a protein which is a potent inhibitor of skeletal myoblast differentiation in vitro . Using ion exchange and molecular exclusion chromatography, we have prepared this protein, which we designate "differentiation inhibitor" (DI), from the materials secreted by BRL cells maintained in serum-free medium . It is a relatively heat-stable protein which is inactivated by treatment with trypsin and mercaptoethanol and has an apparent molecular weight in the range 30,000-36,000 . It exhibits no detectable mitogenic or lectin activity and differs from previously reported inhibitors of myoblast differentiation in several respects . It is active in all skeletal myoblast systems tested (Yaffe's L6 line as well as primary cultures of rat, chick, and Japanese quail myoblasts), and it blocks fusion, elevation of creatine kinase, and increased binding of a-bungarotoxin . Parallel fractionation of fetal bovine serum (FBS) and chick embryo extract (CEE) yields a peak of activity which similarly inhibits myoblast differentiation. We suggest that the differentiation inhibitor from BRL cells may correspond to the differentiation-inhibiting component(s) of FBS and CEE, and we call attention to the possibility that such a substance could play a role in embryonic growth of myoblasts and in satellite cell formation .The mechanisms and control of muscle cell differentiation have been matters of active interest since the feasibility of measuring this process in cultured cells was demonstrated two decades ago . Two views have emerged : in essence, one states that the process is reversibly controlled by components of the medium without any requirement for DNA synthesis (1), and the other proposes that an essential event occurs during a "quantal mitosis" which gives at least one daughter cell irreversibly committed to differentiate (2) . Subsequent papers have reported data interpreted as favoring one or the other view, but in many cases the interpretation has been clouded by the complex nature of uncharacterized medium components such as embryo extract (3) . It seems clear that manipulation of the medium can substantially alter the time course of myoblast differentiation, but the point in the cell cycle at which myoblasts become irreversibly committed to differentiation rather than proliferation remains a matter of disagreement. This situation might be clarified if the processes involved were studied using defined medium components and cloned cells rather than the complex mixtures used in most of the published experiments .A recent report from this laboratory (4) documents the stimulation by the somatomedins of differentiation in cloned L6 myoblasts grown in serum-free medium . Insulin at high