The objective of this study was to examine the association between caffeine intake during pregnancy and fetal mortality in Montevideo, the capital city of Uruguay, taking into account several potential confounding factors. A population-based case-control study was conducted between 1 August 2002 and 31 December 2003. A total of 382 cases and 792 controls were recruited. Cases consisted of women hospitalised with a medically confirmed diagnosis of spontaneous antepartum fetal death, in all maternity hospitals during the study period. Antepartum fetal death was defined as a fetal death in which the attending doctor certified that the death occurred prior to the onset of labour. Fetal deaths were included if they were of at least 20 weeks' gestational age or weighed >350 g. Controls were women who had a live, vigorous and term adequate-for-gestational-age newborn. Multiple gestations and fetuses/newborns with evident congenital malformations were excluded. Only a small proportion of the mothers (8.1% of the cases and 9.5% of the controls) did not consume caffeine during pregnancy. Among consumers, mate drinking was the most frequent source of caffeine in both cases and controls. After controlling for mother's and her partner's education, history of abortions and/or fetal deaths, vomiting/nausea during the first trimester of gestation and attendance for prenatal care, the category of mean caffeine intake of > or = 300 mg/day showed a significantly increased risk of fetal death (OR 2.33 [1.23; 4.41]) compared with no caffeine consumption during pregnancy. The study also found that less-educated women, mothers who did not attend for prenatal care and women with a history of abortions and fetal death were at an increased risk of fetal death. As mate drinking is highly consumed among pregnant women in Uruguay, the association found with fetal death makes it a preventable risk factor.