NONLIPID granular cell tumors are usually considered to be myoblastomas, and in the following review this term will be used freely. However, identity of all these tumors is still open to question. We present 3 cases exemplifying this problem. In 1926 Abrikossoff1 reported a tumor of myoblasts occurring chiefly in relation to striated muscle. He thought it might be due to degenerative lesions following injury or inflammation. In 19312 he further elaborated his ideas on myoblastic myomas, as he called them, feeling that perhaps they were made up of embryonal elements—primitive myoblasts. He divided them into four different types, of which the first three are all granular cell and benign: (1) round, egg-shaped or elongated myoblasts, 20 to 25 microns in length, showing granules but no longitudinal or cross striations; (2) growths in which some of the cells may show longitudinal or cross striations, and (3)