In radiotherapy of pelvic cancers, the X-ray dose to be delivered to the tumour is limited by the tolerance of healthy surrounding tissue. In the past, with orthovoltage equipment, skin tolerance was the main limiting factor. With the introduction of megavoltage equipment, it became possible to deliver a higher radiation dose to deep-seated lesions, but inevitably also to neighbouring normal tissues. As a result, local cure rates were increased, but so was the number of complications. In recent years we encountered a number of serious complications of irradiation of pelvic organs. It was in particular in cases of ovarian cancer where a higher dose of irradiation was applied that these distressing complications were frequent. This experience led to a more conservative irradiation protocol in these cases, resulting in a marked fall in the number of complications. In cervical and endometrial cancer the damage never was as extensive as in the ovarian cancer group. At present many patients who have to ondergo pelvic irradiation, however, will have transient or longer lasting symptoms caused by the irradiation. Modern radiotherapy necessitates the acceptance of a calculated risk of complications in order to achieve a better cure rate. To calculate these risks, one has to know the radiation dose-effect relationship of normal tissues. In quantitative terms these risks are insufficiently known. The normal tissues most at risk when treating pelvic tumours are the bladder, the ureters, the rectum, the sigmoid colon and the small intestine. In this study we limit ourselves to the bowel. The literature regarding postirradiation bowel complications is very confusing. In the first place no two authors used the same criteria for what they consider to be bowel complications. Some authors only include severe complications like stenoses of the bowel and fistulas, in particular those that require surgical intervention (5, 8). Other authors include patients with diarrhea and/ or malabsorption, whereas a third group will include ill defined situations that necessitate long-term hospitalization. It is clear that in this way severe and mild complications are often mixed. The result is that in different series the incidence of severe gastro-intestinal complications varies as much as from I to 15"7o (l-12).lt is to be noted that in these studies damage to the small intestine form only a minority. Another drawback of the reported studies is that all were retrospective. In addition when patients appear to have a recurrence of their tumour in parallel with radiation damage, they are often excluded from further evaluation. In view of the rarity of the severe postirradiation bowel syndromes, a prospective study of these complications would require a very large number of patients. This type of study has never been undertaken.