This paper summarizes the hypothesis proposed to explain the mechanism for AChR localization at the neuromuscular synapse. Two theories have been proposed to explain the neuronal control of extrajunctional AChR. One theory claimed that motoneurons decreased the ACh sensitivity of the extrajunctional membrane through neurotrophic influence. However, direct electrical stimulation of denervated muscles resulted in a decrease of extrajunctional ACh sensitivity, supporting the other hypothesis that loss of extrajunctional AChR of the innervated muscle is directly related to muscle activity per se. AChR clusters (high density of AChR) at the neuro-muscular junction were supposed to result from the association of nerves with preexisting AChR clusters. However, Xenopus nerve-muscle cocultures clearly demonstrated that AChR clusters at the neuromuscular junction were formed after the nerve came in contact with the muscle membrane. Two hypothesis are proposed for nerve-induced formation of AChR clusters. Preferential insertion of AChR into the end-plate was suggested by the finding that AChR messenger RNA was more abundant near to than far from the end-plate in adult muscle fibers. On the other hand, in cultured and embryonic muscles, AChR clusters were formed at nerve-muscle junctions through receptor redistribution which was mediated by the passive diffusion-trap mechanism.