The objective of this work was to identify and evaluate risk factors for shoulder dystocia and for brachial plexus injury in a population-based study. In all, 1,397 parturients with shoulder dystocia occurring in Sweden between 1987 and 1996 were identified among 1,076,545 deliveries using information stored in the Medical Birth Registry (MBR) of the National Board of Health and Welfare, Stockholm. Information is missing for about 1.0% of singleton deliveries in the MBR. Validation of the diagnosis ‘shoulder dystocia’ in the MBR was performed using data of 4 of 63 delivery units and was confirmed in 96.5% of the original medical records. A total of 368 infants had brachial plexus injury (26.3%). However, a considerable variation in the rate of recorded shoulder dystocia was evident when comparing the 63 delivery units throughout Sweden. Overall, the incidence was 1.3 per 1,000 deliveries. A close association was found between birth weight and shoulder dystocia, with an odds ratio (OR) over 30 for a birth weight ≧5,000 g. In the presence of maternal diabetes mellitus or gestational diabetes, the OR for shoulder dystocia was significantly increased in newborns weighing ≧4,000 g. The overall perinatal mortality because of shoulder dystocia was 1.2%. This was increased to 6.4% if the mother had diabetes mellitus. About 25% of all infants suffered a brachial plexus injury. This potentially serious injury even occurred in connection with birth by cesarean section, and an OR for plexus injury over unity was seen in 7 of 9 weight groups; a significantly increased OR was seen for birth weights ≧4,500 g. The birth weight is a strong risk factor for shoulder dystocia, as is diabetes mellitus. A considerable variation in the rate of shoulder dystocia was observed among the different delivery units, probably reflecting difficulties in definition. Brachial plexus injury is observed even after cesarean section, especially if the birth weight is ≧4,500 g.