We compared the effectiveness of systemic corticosteroids with the use of high-dose intravenous immunoglobulin (IVIG) in the treatment of Graves' ophthalmopathy. This was performed as a prospective, nonrandomized study including a blinded ophthalmological and orbital computed tomographic (CT) evaluation. The two groups of patients were not significantly different in relation to sex composition, age distribution, duration of Graves' disease, and ophthalmopathy and previous hyperthyroidism. All patients were followed up by endocrinologic evaluation and blinded ophthalmological (before therapy = B, at the end of therapy = E, and 6 months after the end = 6M) and orbital CT (B and E) evaluations. Twenty-seven patients treated with IVIG were followed up after the end of treatment for an average of 21 months (range 12 to 48 months). Soft tissue involvement (NOSPECS) improved or disappeared in 32 of 35 (90%) patients treated with IVIG and in 25 of 27 (92.5%) patients treated with corticosteroids. Diplopia improved or disappeared in 22 of 29 (75%) patients treated with IVIG and in 16 of 20 (80%) patients treated with corticosteroids. The results observed by clinical evaluation were confirmed with orbital CT score in 30 IVIG patients and in the corticosteroid-treated patients; a significant reduction of extraocular muscle thickness was observed after treatment in both groups. Proptosis improved or disappeared in 20 of 31 (65%) patients treated with IVIG and in 15 of 24 (62%) patients treated with corticosteroids. Mean values of proptosis evaluated by Hertel's exophthalmometer showed a slight reduction both in IVIG as well as in corticosteroid-treated patients. It is interesting to observe that in 28 IVIG-treated patients in whom it was possible to evaluate soft tissue involvement, proptosis and diplopia in the period between the fifth and sixth month from the start of therapy, the most important part of the amelioration (if responders) was already obtained at that time. Responder patients were defined in relation to the decrease in the highest NOSPECS class or grade. Among IVIG-treated patients 26 of 34 (76%) responded; while in the corticosteroid group 18 of 27 (66%) responded to treatment. The prevalences of patients who responded to the treatments were not significantly different in the two groups (Chi-square). The initial values of the subjective eye score were similar in the two groups, and a significant reduction was observed in both. Major side effects requiring discontinuation of the corticosteroid therapy were observed in two patients with hemorrhagic gastritis and in one patient with manic-depressive psychosis. Among 15 patients submitted to the evaluation of bone mineral content before and after corti-costeroid therapy, 4 presented signs of osteoporosis and 3 a reduction of bone mineral content. Moderate and minor side effects were more frequently noted in steroid-treated patients than in the IVIG group. These data suggest that IVIG is safe and effective in reducing the eye changes in patients with Graves' ophthalmopathy.