Objective. To assess the birth of twins in rural Burkina Faso with respect to prevalence, mortality, and provision of obstetric care. Methods. A descriptive demographic analysis of population-based and health service data was performed with multivariate modeling by Poisson regression. Health service data were derived from a document review in health facilities in the Nouna District for the years 1994–2001. The district hospital provides obstetric care for the entire district with its 252,000 population (1998). Population-based data (deliveries, live births, deaths) were derived from the demographic surveillance system in the surveillance area within the Nouna District, Burkina Faso (population under surveillance: 30,988 in 1998). All births in the period from January 1, 1993 to December 31, 1998 (9,457 recorded deliveries with 9,610 live births) were included and followed up until April 30, 2002. The main outcome measures were the prevalence of twin delivery as well as age- and sex-specific mortality risks. Results. Mortality of twins is high, with one out of three dying before reaching the age of five years – 2.5 times the mortality risk of singletons. Among twins, mortality is particularly high in the neonatal period (RR 5.16; CI: 3.6–7.5) and in twins born to mothers above the age of 35 (RR 5.12; CI: 3.5–7.6). The overall population-based prevalence of twin delivery is 1.6% (CI: 1.4–1.9) versus a hospital-based prevalence of 2.8% (CI: 2.2–3.1). Despite this moderate trend towards hospital-based obstetric care, most twins (90.5%) are still delivered outside a hospital setting. Conclusions. The high neonatal mortality in twins points to the need for special care in pregnancy, child birth, and postpartum. Maternity services fail to adequately cover the vast majority of multiple pregnancies and deliveries.