The possibility of a topographic relationship between the spinal cord and the longitudinal axis of a muscle has been explored in two mammalian hamstring muscles: the rat semitendinosus (ST) and biceps femoris (BF). In both muscles the fibers did not extend the full length of the respective muscle bellies but were arranged in longitudinal arrays. There were two such arrays in BF and three in ST; monopolar recordings revealed that each array had a transverse band of endplates extending across the middle part. By stimulating ventral nerve roots in the lumbosacral outflow, it was found that L5 made the greatest contribution to the innervation of both ST and BF, with lesser inputs coming from adjacent roots. In neither ST or BF was there any evidence of a topographic relationship between the spinal cord and the muscle belly.