To analyze the benefits and risks of three common treatments, we randomly assigned 179 patients with Graves' hyperthyroidism as follows: 60 patients, 20-34 yr of age (young adults), received antithyroid drugs for 18 months (medical) or subtotal thyroidectomy (surgical), and 119 patients, 35-55 yr of age (old adults), received medical, surgical, or radioiodine (iodine-131) treatment. The follow-up time was at least 48 months. Antithyroid drugs, surgery, or iodine-131 treatment normalized the mean serum hormone levels within 6 weeks. The risk of relapse was highest in the medically treated young and old adults (42% vs. 34%), followed by that in those treated with iodine-131 (21%) and that in the surgically treated young and old adults (3% vs 8%), respectively. Elevated TSH receptor antibodies at the end of medical therapy or increasing TSH receptor antibodies values after medical or surgical treatment increased the probability of relapse. Development or worsening of ophthalmopathy was not associated with relapse per se. Ninety percent of the subjects in all groups were satisfied with the treatment they received. No significant difference in sick-leave due to Graves' or other diseases was seen during the first 2 yr after initiation of therapy. The increased risk of ophthalmopathy in patients with high serum T3 levels, especially when treated with iodine-131, and the relatively high frequency of relapse after treatment with antithyroid drugs are important factors to consider when selecting therapy for Graves' disease.