A study was made of the effect of cyclosporin A on intimal hyperplasia in an experimental model of arterial autograft. Fifty female Sprague-Dawley rats weighing 250-300 g were employed. Using a microsurgical technique, an arterial autograft measuring approximately 5 mm in length was implanted in right common iliac artery. Two groups were established: group I (control), consisting of 25 animals subjected only to arterial autograft, and group II (preoperative cyclosporin A), also consisting of 25 animals, which received a daily subcutaneous dose of 5 mg/kg cyclosporin A (Sandimmun, Sandoz) for 4 days before the surgical procedure. The animals were sacrificed on postoperative days 7, 14, 21, 30 or 50. Specimens were studied by optical microscopy, transmission and scanning electron microscopy, autoradiography, and morphometry. Endothelialization of the graft zone was slow in the cyclosporin-treated group. Hyperplasia was delayed notably, but at 30 days the hyperplastic process had improved and at 50 days it was similar to that of the control group. In the cyclosporin-treated group, thymidine was not taken up by the medial layer; the absence of medial thymidine uptake correlated with ultrastructural evidence of medial degeneration with lipid vacuolization of the smooth muscle cells and the presence of macrophages. These results suggest that cyclosporin A does not inhibit intimal hyperplasia but instead delays its occurrence, probably because of the drug's toxicity for smooth muscle cells.