Myostatin is a member of the TGF-β superfamily that is essential for proper regulation of skeletal muscle growth. As do other TGF-β superfamily members, myostatin signals into the cell via a receptor complex that consists of two distinct transmembrane proteins, known as the type I and type II receptors. Vertebrates have seven distinct type I receptors, each of which can mix and match with one of five type I receptors to mediate signals for all the TGF-β family ligands. Accumulating evidence indicates that myostatin shares its pair of receptors with activin, and therefore, the question arises about how specificity in signaling is achieved. Our hypothesis is that a mechanism has to exist to restrict myostatin actions to the muscle cells. To investigate this possibility, we compared the effect of endogenous myostatin (myostatin overexpressed by myoblasts) and exogenous myostatin (recombinant myostatin added to the culture medium) in cultured myoblasts. As opposed to exogenous myostatin, endogenous myostatin induced the transcription of a reporter vector in cultured myoblasts. Notably, the myostatin concentrations that failed to induce a response in myoblasts were effective in MCF-7 cells (human mammary carcinoma) and in HepG2 cells (human hepatic carcinoma). Based on our observations, we propose that a mechanism exists that differentially regulates the bioavailability of endogenous and exogenous myostatin to muscle cells. This is consistent with a model in which myostatin actions are exerted in vivo in an autocrine fashion.