In order to assess the effect of propylthiouracil (PTU) or methimazole (MMI) pretreatment on patient outcome after radioiodine therapy, we examined 100 patients with Graves' disease 3, 6, 9, and 12 months after administration of a 10-mCi standard single dose of 131I. They were assigned to one of three groups: no drug (ND) treatment (30 cases); MMI (45 cases); and PTU (25 cases). Antithyroid drugs (ATD) were withdrawn 15 days before radioiodine administration. The groups were similar concerning age, gender, ATD pretreatment duration, goiter size, and initial serum triiodothyronine (T3), thyroxine (T4), free thyroxine (FT4), antithyroid autoantibody levels, 24-hour radioiodine uptake and 131I dose administered per gram of thyroid tissue. ND and MMI groups presented a similar rate of cure of 73.3% and 77.8% respectively (p = NS). In contrast, the PTU group showed a rate of cure of only 32% (p < 0.05). Logistic regression analysis indicated that PTU administration (p = 0.003) and thyroid size (p = 0.02) were the variables related to radioiodine therapy failure. Our data demonstrate that the chance of 131I treatment failure is higher in individuals using PTU than in patients using MMI or not using any ATD before radioiodine (odds ratio [OR] 5.84; 95% confidence interval [CI] 1.82-18.76) suggesting that PTU should be avoided in the treatment of patients with Graves' disease.