Objective: To identify factors associated with the development of neonatal injury in the setting of shoulder dystocia., Study Design: Medical record ICD-9 codes and a computerized perinatal database were reviewed to identify cases of shoulder dystocia from January 1996 to January 2001 in a tertiary care center. For confirmation of the diagnosis and collection of data, both maternal and neonatal charts were then reviewed and neonatal injuries categorized as either neurological (brachial plexus injury) or skeletal (clavicular fracture, humeral fracture). Shoulder dystocia cases were divided into groups based on the presence of neonatal injury at delivery or at discharge (with or without Erb's palsy). The group with neonatal injury was compared for demographic and obstetrical factors to the group without injury (control). chi (2) test, Mann-Whitney test and logistic regression were used as appropriate., Results: During this 5-year period, there were 25,995 deliveries and 206 (0.8%) confirmed cases of shoulder dystocia. Of these cases, 36 (17.5%) had neonatal injury diagnosed at delivery and 25 (12%) remained with significant residual injury at discharge. Of these there were 19 cases of Erb's palsy and six cases of clavicular fracture. No association was found between neonatal injury and maternal age, ethnicity, diabetes, operative vaginal delivery or number of obstetrical maneuvers. However, maternal body mass index >30 kg/m2, a second stage of labor >20 min and a birth weight >4500 g were all associated with an increased risk of neonatal injury at delivery and at discharge, including Erb's palsy. After logistic regression analysis, only a second stage of delivery >20 min remained significantly associated with neonatal injury at discharge., Conclusion: In our population, maternal obesity was associated with an increased risk of neonatal injury after shoulder dystocia. In addition, a short second stage of labor (<20 min) was associated with a lower rate of neonatal injury.