A nested case-control hospital study and a midwife-based community cohort study were conducted in Central Sudan during 1989 and 1990 to assess the contribution of mesoendemic malaria to low birth weight. Malarial infection was determined by maternal history, parasitology, and histopathology. There were significant associations between a maternal history of malaria and low birth weight in the hospital study (adjusted odds ratio (OR) = 1.6, 95% confidence interval (CI) 1.2-2.1) and the community study (OR = 1.7, 95% CI 1.3-2.3). Attributable risk percentages were high and were comparable in the hospital study (22.2%) and the community study (24.5%). A significant trend of increased risk of low birth weight was observed with increasing number of reported malaria attacks, with attacks occurring earlier in pregnancy, and with higher parasitemia. In addition, the risk of low birth weight associated with malaria was higher among primiparous women than among multiparous women. The mean birth weight of infants whose mothers had malaria during pregnancy was significantly lower than the mean birth weight of infants whose mothers did not. Malaria treatment, chemoprophylaxis, and use of insecticides decreased the risk of low birth weight and are recommended as appropriate interventions. These measures should target primigravid women and should be initiated early in pregnancy.