A large, randomized controlled trial to investigate the impact of insecticide-treated curtains (ITC) on child mortality was conducted in an area of seasonal, holoendemic malaria in Burkina Faso. 158 communities totalling some 90,000 people were censused and grouped into 16 geographical clusters, 8 of which were randomly selected to receive ITC in June-July 1994, just prior to the rainy season. In September-October 1995, at the peak period of malaria transmission, a cross-sectional survey was conducted in 84 of the villages. A random sample of 905 children aged 6-59 months was identified and visited. 763 children (84%) were present at the time of the visit and recruited into the study. Mothers were asked about fever in the past 24 h, the child's temperature was taken, and a sample of blood collected to identify and quantify malaria infections and to measure haemoglobin (Hb) levels. Children protected by ITC were less likely to be infected with Plasmodium falciparum (risk ratio = 0.92; 95% CI 0.86, 0.98) or P. malariae (risk ratio = 0.42, 95% CI 0.19, 0.95). The mean intensity of P. falciparum infections was lower among children protected by ITC (899 vs. 1583 trophozoites/microliter; P < 0.001), while the mean Hb level was 0.4 g/dl higher (P < 0.001). While we found no evidence that ITC had an impact on the prevalence of malaria-associated fever episodes, the confidence intervals around our estimates of the impact of ITC on malaria morbidity were wide. We conclude that widespread implementation of ITC in this area of high malaria transmission led to a modest reduction in the prevalence of malaria infection and to a more substantial reduction in the intensity of these infections which caused increased Hb levels. We were unable to demonstrate any impact of ITC on malaria morbidity, but the wide confidence intervals around our point estimates do not preclude the possibility of a substantial impact.