We evaluated the outcome of 131I treatment of Graves' disease in two different protocols (old and new protocol) of adjusted accumulation dose from 1988 to 1995. Adjusted accumulation doses of patients with above 50 g thyroid weights were increased by 5-20 Gy/g tissue in new protocol compared to those in old one. In 166 patients treated with single and plural doses of 131I treatment in 1990 (Group In), the therapeutic doses were calculated according to new protocol and in 130 patients in 1988 (Group Io), according to old one, modification of Quimby's formula. The patients treated with plural doses were classified as hyperthyroidism because the efficacies of the first treatments with 131I were insufficient. At the 5-yr follow up, the incidence of hypothyroid in Group In was 9%, subclinical hypothyroid 17%, euthyroid 30%, subclinical hyperthyroid 7%, hyperthyroid 37%. In Group Io, 11% of the patients were hypothyroid, 6% subclinical hypothyroid, 29% euthyroid, 3% subclinical hyperthyroid, 51% hyperthyroid. The incidence of hyperthyroid in Group In was lower than that in Group Io (p < 0.05). There were no significant differences in hypothyroid and euthyroid. In conclusion, we suggest that an adjusted dose of 131I in relation to the patients' thyroid weight shows some room for improvement.