Treatment of Graves' disease with iodine-131 (131I) is well-known; however, all patients do not respond to a single dose of 131I and may require higher and repeated doses. This study was carried out to identify the factors, which can predict treatment failure to a single dose of 131I treatment in these patients. Data of 150 patients with Graves' disease treated with 259-370 MBq of 131I followed-up for at least 1-year were retrospectively analyzed. Logistic regression analysis was used to predict factors which can predict treatment failure, such as age, sex, duration of disease, grade of goiter, duration of treatment with anti-thyroid drugs, mean dosage of anti-thyroid drugs used, 99mTc-pertechnetate (99mTcO4-) uptake at 20 min, dose of 131I administered, total triiodothyronine and thyroxine levels. Of the 150 patients, 25 patients required retreatment within 1 year of initial treatment with 131I. Logistic regression analysis revealed that male sex and 99mTcO4- uptake were associated with treatment failure. On receiver operating characteristic (ROC) curve analysis, area under the curve (AUC) was significant for 99mTcO4- uptake predicting treatment failure (AUC = 0.623; P = 0.039). Optimum cutoff for 99mTcO4- uptake was 17.75 with a sensitivity of 68% and specificity of 66% to predict treatment failure. Patients with >17.75% 99mTcO4- uptake had odds ratio of 3.14 (P = 0.014) for treatment failure and male patients had odds ratio of 1.783 for treatment failure. Our results suggest that male patients and patients with high pre-treatment 99mTcO4- uptake are more likely to require repeated doses of 131I to achieve complete remission.