This Phase II protocol was designed around the format of a previously reported study for the palliation of advanced pelvic malignancies (RTOG 7905). The large dose per fraction (1000 cGy) of the first study unexpectedly demonstrated frequent late gastrointestinal toxicity and was replaced in the present study by 2 days of twice daily fractionation (370 cGy/fraction totaling 1480 cGy/course) based on linear quadratic consideration of acute and late toxicities. The interval between courses of 4 weeks was retained and the total dose for three courses was 4440 cGy. A total of 152 patients were entered of which 142 have sufficient follow-up information for analysis. Fifty-nine percent of the patients completed all three courses, 20% completed two courses, and 20% received only one course. The primary sites were: gynecological (39.4%); colorectal (32.4%); genitourinary (24.7%); and other (2.8%). The best overall response was: complete remission (10%); partial remission (22%); no change (24%); progression (10%); and unknown (27%). For the patients completing all three courses, the response was: complete remission (14%); partial remission (31%); no change (40%); progression (7%); and unknown (8%). Median survival was 4.5 months and the actuarial survival at 12 months is 19%. In RTOG 7905, 90% of the late toxicities appeared by 12 months. For RTOG 8502, there were 32 patients alive at 9 months and 19 patients alive at 12 months available for evaluation of late toxicity. There has been one late grade 3 GI toxicity reported and only one acute grade 3 gastrointestinal toxicity. Even if the true rate of late toxicity for 8502 were 20%, the chance of seeing none or one toxicity would be 0.007. This toxicity report represents a marked reduction of the grade 3 and 4 late toxicity seen in RTOG 7905 (37% at 9 months and 45% at 12 months) without lowering the tumor response rate. The interval between courses in both protocols allows for potential tumor proliferation. To test for the effect of tumor proliferation, RTOG 8502 is continuing as a Phase III randomization between 4 weeks and 2 weeks separation.