Résumé : BACKGROUND: Comparing two surveys performed in Bolivia in 1992-1994 and 1999-2001, we reported a significant decrease in the proportions of severe and mortal forms of congenital Chagas disease. This might be due to a reduction of vectorial density (VD) in maternal residence area, raising the question of a possible causal relationship between such VD, maternal parasitaemia and prognosis of congenital infection with Trypanosoma cruzi. METHOD: Comparisons of haematological and parasitological data obtained from Bolivian mothers infected with T. cruzi, and of clinical and biological data obtained from their infected and uninfected newborns, stratified according to VD in the area of maternal residence. RESULTS: i) Blood hematocrit rates or hemoglobin amounts were within the normal ranges and similar in all the maternal groups, whatever the VD in their areas of residence; ii) mothers living in high VD areas displayed a higher frequency of hemocultures positive for T. cruzi; iii) newborns congenitally infected with T. cruzi, but not uninfected babies born from infected mothers, displayed higher frequencies of very low Apgar scores, low birth weights, prematurity, respiratory distress syndrome or anasarca, as well as higher mortality rates when their mothers lived in areas of high VD. CONCLUSION: Frequent bites of blood sucking Reduvidae during pregnancy do not induce maternal anaemia, but, likely through multiple maternal re-infections with T. cruzi, increase maternal parasitemia and worsen congenital Chagas disease. Maternal dwelling in areas of high VD is associated with a serious increased risk of severe and mortal congenital Chagas disease.