Our objectives were to analyze the risks versus benefits of flexible sigmoidoscopy to the pregnant female and fetus. We retrospectively studied 24 consecutive pregnant patients admitted to four university hospitals during seven years who underwent 26 flexible sigmoidoscopies. Sigmoidoscopy indications included hematochezia in 11, diarrhea in 12, abdominal pain in 7, constipation in 2, and occult rectal bleeding in 1. Seven patients were in the first trimester of pregnancy, nine were in the second trimester, and eight were in the third trimester. Sigmoidoscopy provided helpful clinical information in all patients. Twelve patients had a lesion diagnosed by sigmoidoscopy, including reactivation of Crohn's colitis, reactivation of ulcerative colitis, infectious colitis, nonspecific colitis, bleeding internal hemorrhoids, pseudomembranous colitis, anastomotic ulcer, and newly diagnosed Crohn's colitis. In particular, nine of 11 patients with rectal bleeding had a lesion identified by sigmoidoscopy. No endoscopic complications occurred to any pregnant female. Two pregnant patients underwent repeat sigmoidoscopy without complications. Fetal outcome was ascertained in all but one pregnancy. Eighteen pregnant females delivered healthy infants (16 at full term, two at 35 or 36 weeks). Their mean Apgar scores were 8.8 +/- 0.4 SD at 1 min, and 9.0 +/- 0.4 SD at 5 min. One diabetic and hypertensive female suffered an involuntary abortion nine weeks after sigmoidoscopy, which appeared unrelated to the sigmoidoscopy. Four pregnancies were voluntarily aborted. This study suggests that flexible sigmoidoscopy does not induce labor or result in congenital malformations, that sigmoidoscopy is not contraindicated during pregnancy, and that sigmoidoscopy should be considered in medically stable pregnant patients with significant gastrointestinal bleeding.(ABSTRACT TRUNCATED AT 250 WORDS)