The aim of this study was to attempt to identify factors predictive of uterine rupture following operative hysteroscopy. A MEDLINE and EMBASE search (1980 to April 2006) using the keywords "hysteroscopy", "uterine rupture" and "pregnancy" was performed with no limitations of languages. 18 cases of uterine rupture following operative hysteroscopy were retrieved. Hysteroscopic metroplasty (uterine septa and synechiae) was involved in 16 (89%) cases. Hysteroscopic resection was performed with monopolar current cutting, rigid scissors and laser in respectively 14 (78%), 3 (16.5%) and 1 (5.5%) cases. Uterine perforation occurred in 10 (55.5%) cases. The interval between hysteroscopy and subsequent pregnancies varied from 1 month to 5 years with an average delay of 16 months. Uterine rupture occurred between 19 and 41 weeks, with no labour in 12 (66.5%) cases. Four fetal and one maternal deaths followed uterine rupture. Hysteroscopic metroplasty may expose patients to an increased risk of uterine rupture in subsequent pregnancy. Uterine perforation and/or the use of current monopolar section during operative hysteroscopy increase this risk. Uncomplicated hysteroscopic resection of submucous myomas and endometrial polyps do not alter obstetrical outcome. Considering hysteroscopic metroplasty, the use of coaxial bipolar electrode should be preferred.