A randomized clinical trial was performed to clarify whether pretreatment with propylthiouracil (PTU) before radioiodine ((131)I) therapy influences the final outcome of this therapy, as has been indicated by retrospective studies. Untreated consecutive hyperthyroid patients with Graves' disease (n = 23) or a toxic nodular goiter (n = 57) were randomized to either PTU (+PTU; n = 39) or no pretreatment (-PTU; n = 41) before compensated (131)I therapy. The median PTU dose was 100 mg, which was discontinued 4 d before treatment. The median (131)I activity was 302 MBq (range, 87-600 MBq). After (131)I therapy, the serum free T(4) index increased in the +PTU group from 97.7 +/- 47.5(+/-sd) nmol/liter at the time of therapy to 152.3 +/- 77.6 nmol/liter at 3 wk (P < 0.001) and 140.4 +/- 75.9 nmol/liter at 6 wk (P < 0.001). In the -PTU group, the serum free T(4) index, which was initially 254.3 +/- 145.7 nmol/liter, decreased significantly to 212.0 +/- 113.0 nmol/liter at 3 wk (P < 0.05) and 165.8 +/- 110.0 nmol/liter at 6 wk (P < 0.005). After 1 yr of follow-up, the treatment failure rate in patients with a toxic nodular goiter was four times higher in the +PTU group than in the -PTU group (nine of 20 vs. three of 25 patients; P = 0.06), whereas the difference among patients with Graves' disease was less obvious (four of six vs. four of nine; P = 0.81). Patients in the +PTU group who were cured had higher serum TSH (s-TSH) levels at the time of (131)I therapy than those who were not cured. By adjusting for a possible interfactorial relationship through a regression analysis, including the s-TSH level and type of disease, only PTU pretreatment had a significant adverse effect on the cure rate (P = 0.03). In conclusion, this randomized trial demonstrates that PTU pretreatment reduces the cure rate of (131)I therapy in hyperthyroid diseases, although this adverse effect seems to be attenuated by the concomitant rise in s-TSH.