Primary cultures prepared from dermal and plexiform neurofibromas contain Schwann-like cells and fibroblast-like cells. SLC are elongated and bipolar or multipolar. By indirect immunofluorescence light microscopy, living SLC bind antibodies against laminin and against nerve growth factor receptor to their surface, but not antibodies against fibronectin. In these respects, cultured SLC are indistinguishable from cultured human adult Schwann cells. FLC are flat and pleomorphic. By indirect immunofluorescence light microscopy, living FLC bind antibodies against fibronectin but not against laminin or NGFR. In these respects, cultured FLC are indistinguishable from cultured human adult endoneurial fibroblasts. Considerable purification of viable SLC from SLC/FLC mixed cultures can be achieved by flow cytofluorometry using a monoclonal anti-NGFR antibody. Tritiated thymidine radioautography indicated that mitosis of SLC in mixed SLC/FLC cultures prepared from dermal neurofibromas is infrequent in MEM with 10% calf serum, more frequent in RPMI 1640 medium with 15% fetal calf serum. Central nervous system axolemmal fragments (rat or human) elicited a greater than 10-fold SLC proliferative response in mixed SLC/FLC cultures from three of seven dermal neurofibromas (from six patients with neurofibromatosis), but had no effect on SLC mitosis in cultures from the other four dermal neurofibromas. SLC mitosis was inhibited by concentrations of cyclic adenosine 3',5'-monophosphate analogues known to stimulate proliferation of normal rat Schwann cells. Glial growth factor partially purified from bovine pituitaries stimulated SLC mitosis both in SLC/FLC mixed cultures and in cultures of purified SLC. The studies we have described indicate that neurofibroma SLC can be cultured, unequivocally identified in culture by morphological and immunohistological criteria, purified, and stimulated to proliferate by several Schwann cell mitogens. Further quantitative comparisons of the baseline and mitogen-stimulated rates of proliferation of SLC and age-matched control human Schwann cells are needed, however, to determine which of the two alternate pathogenetic mechanisms for formation of neurofibromas mentioned in the introduction is correct.