- Vacuum extraction, more widely used than forceps for operative vaginal delivery in Europe, is said to have failed when it is unduly difficult or there are more than two cup disengagements. This retrospective study of singleton, vertex deliveries at more than 36 weeks' gestation sought to identify obstetrical risk factors for a failed trial of vacuum extraction. A total of 211 trials taking place in the years 1990-1998 were reviewed; 5.4% of attempts (113) failed, leading to cesarean delivery within 15 minutes. Extraction was done only when the cervix was completely dilated, the membranes had ruptured, and the fetal head was engaged. Traction was done slowly and was stopped if there was any loss of vacuum. Pressure was limited to 0.8 kg/cm 2 . Only two reapplications of the metal cup were permitted. Three instances of lost suction were construed as method failure. Failed vacuum extraction was significantly and linearly associated with high birth weights; these infants were likelier to be large for gestational age and to weigh more than 4000 g. A lack of prenatal care also correlated with a higher rate of failed extraction. On multiple logistic regression analysis, both deficient prenatal care and a birth weight greater than 4000 g were independent risk factors for failed vacuum extraction. Women in whom extraction failed had significantly more cervical and uterine tears than when the procedure succeeded, and the relationship with uterine rupture persisted after adjusting for previous cesarean delivery, Failed extraction correlated with postpartum anemia and also with higher rates of intrapartum and postpartum fetal death (odds ratios: 35.9 for intrapartum fetal death and 9.1 for postpartum death). Both maternal and neonatal outcomes are compromised by a failed trial of vacuum extraction. An estimated fetal weight exceeding 4000 g should be viewed as a relative contraindication, particularly for women who have not received adequate prenatal care.