Sixty-four patients were randomized prior to pelvic radiotherapy into one of three dietary groups: the control group maintained a regular diet except that they drank at least 480 cc of milk daily; the lactose-restricted group was placed on a lactose-restricted diet; and the lactase group drank at least 480 cc of milk with lactase enzyme added to hydrolyze 90% of the lactose. The patients kept records of their stool frequency and the number of diphenoxylate tablets required to control their diarrhea during a 5 week course of standard whole pelvis irradiation. The data does not support the concept that one of the mechanisms of radiation-induced diarrhea associated with pelvic irradiation is a reduction the ability of the intestine to hydrolyze ingested lactose due to the effect of the radiation on the small intestine. There was not a significant difference in stool frequency or diphenoxylate usage among the dietary groups.