BACKGROUND: An increasing number of people survive cancer but a significant proportion have gastrointestinal side effects as a result of radiotherapy (RT), which impairs their quality of life (QoL). OBJECTIVES: To determine which prophylactic interventions reduce the incidence, severity or both of adverse gastrointestinal effects among adults receiving radiotherapy to treat primary pelvic cancers. SEARCH METHODS: We conducted searches of CENTRAL, MEDLINE, and Embase in September 2016 and updated them on 2 November 2017. We also searched clinical trial registries. SELECTION CRITERIA: We included randomised controlled trials (RCTs) of interventions to prevent adverse gastrointestinal effects of pelvic radiotherapy among adults receiving radiotherapy to treat primary pelvic cancers, including radiotherapy techniques, other aspects of radiotherapy delivery, pharmacological interventions and non-pharmacological interventions. Studies needed a sample size of 20 or more participants and needed to evaluate gastrointestinal toxicity outcomes. We excluded studies that evaluated dosimetric parameters only. We also excluded trials of interventions to treat acute gastrointestinal symptoms, trials of altered fractionation and dose escalation schedules, and trials of pre- versus postoperative radiotherapy regimens, to restrict the vast scope of the review. DATA COLLECTION AND ANALYSIS: We used standard Cochrane methodology. We used the random-effects statistical model for all meta-analyses, and the GRADE system to rate the certainty of the evidence. MAIN RESULTS: We included 92 RCTs involving more than 10,000 men and women undergoing pelvic radiotherapy. Trials involved 44 different interventions, including radiotherapy techniques (11 trials, 4 interventions/comparisons), other aspects of radiotherapy delivery (14 trials, 10 interventions), pharmacological interventions (38 trials, 16 interventions), and non-pharmacological interventions (29 trials, 13 interventions). Most studies (79/92) had design limitations. Thirteen studies had a low risk of bias, 50 studies had an unclear risk of bias and 29 studies had a high risk of bias. Main findings include the following:Radiotherapy techniques: Intensity-modulated radiotherapy (IMRT) versus 3D conformal RT (3DCRT) may reduce acute (risk ratio (RR) 0.48, 95% confidence interval (CI) 0.26 to 0.88; participants = 444; studies = 4; I = 77%; low-certainty evidence) and late gastrointestinal (GI) toxicity grade 2+ (RR 0.37, 95% CI 0.21 to 0.65; participants = 332; studies = 2; I = 0%; low-certainty evidence). Conformal RT (3DCRT or IMRT) versus conventional RT reduces acute GI toxicity grade 2+ (RR 0.57, 95% CI 0.40 to 0.82; participants = 307; studies = 2; I = 0%; high-certainty evidence) and probably leads to less late GI toxicity grade 2+ (RR 0.49, 95% CI 0.22 to 1.09; participants = 517; studies = 3; I = 44%; moderate-certainty evidence). When brachytherapy (BT) is used instead of external beam radiotherapy (EBRT) in early endometrial cancer, evidence indicates that it reduces acute GI toxicity (grade 2+) (RR 0.02, 95% CI 0.00 to 0.18; participants = 423; studies = 1; high-certainty evidence).Other aspects of radiotherapy delivery: There is probably little or no difference in acute GI toxicity grade 2+ with reduced radiation dose volume (RR 1.21, 95% CI 0.81 to 1.81; participants = 211; studies = 1; moderate-certainty evidence) and maybe no difference in late GI toxicity grade 2+ (RR 1.02, 95% CI 0.15 to 6.97; participants = 107; studies = 1; low-certainty evidence). Evening delivery of RT may reduce acute GI toxicity (diarrhoea) grade 2+ during RT compared with morning delivery of RT (RR 0.51, 95% CI 0.34 to 0.76; participants = 294; studies = 2; I = 0%; low-certainty evidence). There may be no difference in acute (RR 2.22, 95% CI 0.62 to 7.93, participants = 110; studies = 1) and late GI toxicity grade 2+ (RR 0.44, 95% CI 0.12 to 1.65; participants = 81; studies = 1) between a bladder volume preparation of 1080 mls and that of 540 mls (low-certainty evidence). Low-certainty evidence on balloon and hydrogel spacers suggests that these interventions for prostate cancer RT may make little or no difference to GI outcomes.Pharmacological interventions: Evidence for any beneficial effects of aminosalicylates, sucralfate, amifostine, corticosteroid enemas, bile acid sequestrants, famotidine and selenium is of a low or very low certainty. However, evidence on certain aminosalicylates (mesalazine, olsalazine), misoprostol suppositories, oral magnesium oxide and octreotide injections suggests that these agents may worsen GI symptoms, such as diarrhoea or rectal bleeding.Non-pharmacological interventions: Low-certainty evidence suggests that protein supplements (RR 0.23, 95% CI 0.07 to 0.74; participants = 74; studies = 1), dietary counselling (RR 0.04, 95% CI 0.00 to 0.60; participants = 74; studies = 1) and probiotics (RR 0.43, 95% CI 0.22 to 0.82; participants = 923; studies = 5; I = 91%) may reduce acute RT-related diarrhoea (grade 2+). Dietary counselling may also reduce diarrhoeal symptoms in the long term (at five years, RR 0.05, 95% CI 0.00 to 0.78; participants = 61; studies = 1). Low-certainty evidence from one study (108 participants) suggests that a high-fibre diet may have a beneficial effect on GI symptoms (mean difference (MD) 6.10, 95% CI 1.71 to 10.49) and quality of life (MD 20.50, 95% CI 9.97 to 31.03) at one year. High-certainty evidence indicates that glutamine supplements do not prevent RT-induced diarrhoea. Evidence on various other non-pharmacological interventions, such as green tea tablets, is lacking.Quality of life was rarely and inconsistently reported across included studies, and the available data were seldom adequate for meta-analysis. AUTHORS' CONCLUSIONS: Conformal radiotherapy techniques are an improvement on older radiotherapy techniques. IMRT may be better than 3DCRT in terms of GI toxicity, but the evidence to support this is uncertain. There is no high-quality evidence to support the use of any other prophylactic intervention evaluated. However, evidence on some potential interventions shows that they probably have no role to play in reducing RT-related GI toxicity. More RCTs are needed for interventions with limited evidence suggesting potential benefits.