Regional differences in stillbirth and neonatal death rates in Sweden were studied and a cause-of-death analysis was done in the 4 counties (among 24) with an increased mortality. The study is based on a computerized evaluation of infant cause of death, using a slightly modified Wigglesworth classification and a hierarchical classification (NICE: Neonatal and Intrauterine death Classification according to Etiology). Differences between the identified counties with respect to specific causes of death were demonstrated. Conclusion: There are differences between Swedish counties with respect to the risk for stillbirth or neonatal death. The NICE cause-of-death classification can be used for the routine surveillance of stillbirths and neonatal deaths in a population and can help in pinpointing weak elements in antenatal, delivery and neonatal care.