RTOG 83-07 is a Phase II randomized protocol designed to compare the efficacy and toxicity of Megestrol vs Diethylstilbestrol (DES) used as cytoreductive agents prior to and during radiotherapy. The end-points of this study include tumor clearance rate, effect on serum testosterone, local-regional control, disease-free interval, and survival. Eligible patients were those with histologically confirmed locally advanced adenocarcinoma, clinical Stage B2 and C without regional lymph node involvement, or with lymph node involvement limited to the pelvis. Patients with medical conditions potentially predisposing to cardiovascular (thromboembolic) sequelae of endocrine therapy were not eligible. Patients were stratified by clinical stage, histological grade, and nodal status and were randomized to receive either Megestrol 40 mg PO tid or Diethylstilbestrol 1 mg PO tid. The drugs were started 2 months prior to initiation of radiotherapy and were continued throughout the radiotherapy course. Radiotherapy consisted of 44 to 46 Gy, 1.8 to 2 Gy per day to the regional lymphatics followed by a boost to the prostate consisting of 20 to 25 Gy, 1.8 to 2 Gy per day to a total of 65 to 70 Gy. Serum testosterone levels were recorded throughout the treatment course. Tumor response was assessed clinically and radiographically (CT scan). From March 1983 through June 1986 a total of 203 patients were accessioned to the study; 197 were analyzable. Correlation of the incidence of drug related toxicity and treatment arm assignment revealed a significantly higher incidence of complications in the Diethylstilbestrol (DES) arm. The most prominent were the differences in the incidence of gynecomastia (55% vs 7%) and fluid retention (21% vs 6%). The incidence of thromboembolic phenomena was comparable (8% vs 5% in the Megestrol arm). Although patients on the DES arm demonstrated a significantly greater median decrease in testosterone level, correlation of the treatment assigned to the rate of tumor regression and the incidence of complete response revealed no significant difference between the arms. At 3 years only 6.5% of the evaluable patients manifested evidence of local failure. The results of the study indicate comparable efficacy (using tumor clearance as an end-point) of DES and Megestrol. While DES appears more effective in suppressing testosterone it is also associated with a higher incidence of toxicity. The cytoreduction (using either DES, Megestrol, or an alternative regimen) concept remains to be tested in a Phase III study comparing it to radiotherapy alone.