OBJECTIVE To identify factors involved in the development of fetal cephalohematoma from vacuum extraction. METHODS Patients at > or = 34 weeks' gestation were randomly assigned to delivery by vacuum (n = 322) using the continuous (n = 164) or intermittent (n = 158) technique. Neonatal outcome with cephalohematoma was analyzed subsequently and related to prospectively recorded data. RESULTS Approximately equal numbers of cephalohematoma were recorded in the two groups (continuous 20, intermittent 17; P = .686). Station at point of application (P = .008), increasing asynclitism (P < .001) and increasing application to delivery time (P = .002) correlated significantly with cephalohematoma. Only the last two factors achieved significance after stepwise multiple logistic regression analysis. Factors that did not achieve statistical significance were gestational age (P = .755), birth weight (P = .982), instrumental rotation (P = .896) and previous vaginal delivery (P = .051). CONCLUSIONS In this prospective, randomized, controlled trial of vacuum-assisted delivery, the only predelivery factor found to predispose to neonatal cephalohematoma formation was increasing asynclitism. Although cephalohematoma formation was more likely to develop as the duration of vacuum application increased during delivery, only 28% of neonates exhibited this finding when the time from vacuum application to delivery exceeded five minutes.