Objectives: To investigate the association between per capita pesticide consumption and infant mortality rates from CNS and CVS congenital malformations in microregions in the South and Southeast Region of Brazil. Material and Methods: An ecological study was conducted using data on pesticide expenditure in 1985 and 1996, and deaths caused by CNS and CVS malformations in infants under 1 year old in 1986-1990 and 1997-2001, respectively. Per capita pesticide consumption and infant mortality rates were calculated for each microregion. Microregions were grouped according to quintiles of pesticide consumption, taking the first quintile as reference. The association between pesticide consumption and infant mortality was examined by calculating Spearman correlation coefficients (r) and mortality rate ratios (RR), stratifying by gender and type of microregion (urban or rural). Results: Significant and positive correlations between per capita pesticide consumption and rates of mortality due to CNS and CVS defects were observed in rural but not urban microregions. In general, mortality RRs for the 2 types of malformations in rural microregions were significantly higher in each quintile of pesticide consumption compared to the lowest quintile in the 2 study periods, with elevations ranging between 10% and 30%. Likewise, mortality RRs in these microregions showed significant trends of increase across quintiles of pesticide consumption in both study periods. In urban areas, however, mortality RRs from both CNS and CVS malformations were weak and not statistically significant, and a trend of increase of mortality with increasing pesticide usage was not observed. Conclusions: The results show the relevance of pesticide exposure in rural areas with intense agricultural activity, suggesting that such prenatal exposures may be related with the occurrence of certain congenital defects.