OBJECTIVE To audit glove perforations at laparotomies for gynaecological cancers. METHODS Gynaecological oncology unit, cancer centre, London. METHODS Prospective audit. METHODS Twenty-nine laparotomies for gynaecological cancers over 3 months. METHODS Gloves used during laparotomies for gynaecological cancer were tested for perforations by the air inflation and water immersion technique. Parameters recorded were: type of procedure, localisation of perforation, type of gloves, seniority of surgeon, operation time and awareness of perforations. METHODS Glove perforation rate. RESULTS Perforations were found in gloves from 27/29 (93%) laparotomies. The perforation rate was 61/462 (13%) per glove. The perforation rate was three times higher when the duration of surgery was more than 5 hours. The perforation rate was 63% for primary surgeons, 54.5% for first assistant, 4.7% for second assistant and 40.5% for scrub nurses. Clinical fellows were at highest risk of injury (94%). Two-thirds of perforations were on the index finger or thumb. The glove on the nondominant hand had perforations in 54% of cases. In 50% of cases, the participants were not aware of the perforations. There were less inner glove perforations in double gloves compared with single gloves (5/139 versus 26/154; P = 0.0004, OR = 5.4, 95% CI 1.9-16.7). The indicator glove system failed to identify holes in 44% of cases. CONCLUSIONS Glove perforations were found in most (93%) laparotomies for gynaecological malignancies. They are most common among clinical fellows, are often unnoticed and often not detected by the indicator glove system.