Nerve segments approximately 7 mm long were excised from the predegenerated sciatic nerves of mice, and treated 5 times by repetitive freezing and thawing to kill the Schwann cells. Such treated nerve segments were grafted into the original places so as to be in contact with the proximal stumps. The animals were sacrificed 1, 2, 3, 5, 7 and 10 days after the grafting. The grafts were examined by electron microscopy in the middle part of the graft, i.e. 3-4 mm distal to the proximal end and/or near the proximal and distal ends of the graft. In other instances, the predegenerated nerve segments were minced with a razor blade after repetitive freezing and thawing. Such minced nerves were placed in contact with the proximal stumps of the same nerves. The animals were sacrificed 10 days after the grafting. Within 1-2 days after grafting, the dead Schwann cells had disintegrated into fragments. They were then gradually phagocytosed by macrophages. The basal laminae of Schwann cells, which were not attacked by macrophages, remained as empty tubes (basal lamina scaffolds). In the grafts we examined, no Schwann cells survived the freezing and thawing process. The regenerating axons always grew out through such basal lamina scaffolds, being in contact with the inner surface of the basal lamina (i.e. the side originally facing the Schwann cell plasma membrane). No axons were found outside of the scaffolds. One to two days after grafting, the regenerating axons were not associated with Schwann cells, but after 5-7 days they were accompanied by Schwann cells which were presumed to be migrating along axons from the proximal stumps. Ten days after grafting, proliferating Schwann cells observed in the middle part of the grafts had begun to sort out axons. In the grafts of minced nerves, the fragmented basal laminae of the Schwann cells re-arranged themselves into thicker strands or small aggregations of basal laminae. The regenerating axons, without exception, attached to one side of such modified basal laminae. Collagen fibrils were in contact with the other side, indicating that these modified basal laminae had the same polarity in terms of cell attachment as seen in the ordinary basal laminae of the scaffolds.(ABSTRACT TRUNCATED AT 400 WORDS)