We have studied various aspects of MAP-1 and MAP-2 from neuronal as well as nonneuronal sources. MAP-1 and MAP-2 polymerized from brain were resolved into a number of subcomponents upon electrophoresis on low percentage gels. Based on peptide mappings performed under a variety of different conditions, we conclude that the three major subcomponents of MAP-1 have very similar, though not identical structures. The two major MAP-2 subcomponents might have identical structure, because their peptide maps were hardly distinguishable. The apparent microheterogeneity of high Mr MAPs is not yet understood on a molecular basis. Proteolysis during isolation or a different degree of phosphorylation, however, seems to be an unlikely cause for microheterogeneity. When localized on microtubules polymerized in vitro by electron microscopy, both MAP-1 and MAP-2 polypeptides apparently form helical arrays on the polymer's surface with periodicities of 100 nm. In the presence of taxol, MAPs form irregular and bulky extensions. Both MAPs are found to be widespread in neuronal as well as nonneuronal cells. MAP-1- and MAP-2-related polypeptides, together with other high Mr proteins, such as plectin, were associated with microtubules polymerized by taxol from extracts of a nonneuronal cultured cell line. MAP-2 from cultured cells was found to be extremely sensitive to proteolysis, in particular in the presence of free Ca-ions. MAP-1 and MAP-2 generally were found associated with typical microtubule structures such as interphase and spindle microtubules and primary cilia. A differential distribution of MAP-1 and MAP-2 was clearly evident in neural tissues, where MAP-2 was restricted to cell bodies and dendrites, whereas MAP-1 was present also in axons. Moreover, a differential distribution of MAPs and tubulin was observed in de-and regenerating peripheral nerve, and in a few occasions, also with nonneuronal cells. A quite unexpected result was the identification of a protein in the extracellular matrix of cultured fibroblast cells, which has antigenic determinants in common with MAP-1 and MAP-2 from brain. As a whole, the data presented support a concept in which a family of structurally homologous, though not identical, high Mr polypeptides constitute the crosslinking elements between microtubules and various other cellular components. The structural diversity of these polypeptides might play a role in the development and dynamic changes in the cytoskeletal architecture.