Rats with chronic streptozotocin-induced diabetes develop enlargement of the alimentary tract, loss of fecal consistency, and autonomic neuropathy involving the extrinsic innervation of the ileum. In this study we have continued the characterization of the unmyelinated axonopathy involving the ileal mesenteric nerves of outbred Sprague-Dawley rats diabetic for 1.5 to 10 months using ultrastructural methods. Axonopathy in the alimentary tract of chronically diabetic rats is characterized by modest to marked dilation of axons by distinctive subcellular organelles identical with those described in experimental and clinical axonal dystrophies. Lesions are infrequent 1.5 and 3 months after induction of diabetes, increasing rapidly in numbers thereafter. Axonal lesions are reproducibly encountered in the ileum, cecum, and rectum but not in the proximal jejunum of chronically diabetic rats. Sectioning of mesenteric nerves along their longitudinal axes demonstrates the focal and in some instances apparent terminal nature of the axonal dilation. Disappearance of dystrophic axons distal (peripheral) to mesenteric crush injury confirms the origin of the axons extrinsic to the gut wall. Only rare examples of dystrophic diabetic axonopathy were detected within the wall of the ileum, presumably representing involvement of terminal axonal segments within the intrinsic ganglia. Examination of the mesenteric nerves of 18- and 30-month-old control rats failed to demonstrate dystrophic axonopathy, providing evidence that diabetic axonopathy does not represent premature development of an aging change in mesenteric nerves. Examination of the autonomic innervation of various other tissues including spleen, bladder, vas deferens, and iris, as well as the phrenic, sciatic, vagus, and tail nerves of 4- to 12-month diabetic animals, failed to demonstrate reproducible axonopathy comparable to that involving the alimentary tract of the same animals. The paravascular fascicles of ileal mesenteric nerves of 6- to 7-month diabetic and age-matched control rats were examined by morphometric methods and failed to demonstrate significant loss of axons or an appreciable shift in mean fiber diameter.