An institution based case-control study to determine risk factors for stillbirths was conducted in the city of Natal, NE Brazil, where 90% of deliveries take place in health facilities. Two hundred thirty-four singleton stillborn cases were compared to 2555 liveborn singleton control infants of normal birth-weight and gestational age. Information was obtained by postnatal interview and anthropometry, and review of medical records. Univariate analyses revealed a large number of potential risk factors, but after adjustment by logistic regression only six factors remained significantly associated with stillbirth. These were low maternal weight, less than or equal to 50 kg and a history of pregnancy loss, both with odds ratios (OR) of 1.8, inadequate prenatal care defined as less than five visits (OR = 1.9), gestational complications (OR = 14.2), intrapartum complications (OR = 2.0), and congenital malformations (OR = 8.7). There was also an increased risk of stillbirth among older mothers who smoked (OR = 1.4), and evidence of an interaction between smoking and complications of pregnancy. From the public health perspective, the most important factors amenable to intervention were inadequate prenatal care and antenatal or intrapartum complications which were associated with substantial attributable risks (23.8%, 35.2%, and 10.2%, respectively). Thus, in this population, future reductions of the high stillbirth rate (27.2 per 1000 births) will largely depend on the coverage, utilization, and quality of antenatal and intrapartum care.