Research has suggested possible human health effects from low-level widespread exposure to environmental contaminants. We employed a novel exposure estimation technique using a publicly available data set to examine atrazine exposure, a suspected endocrine disruptor, in relation to breast cancer risk for women living in rural areas of Wisconsin. Incident breast cancer cases who were 20-79 years of age from 1987 to 2000 (n=3,275) and living in rural areas of Wisconsin at the time of interview were identified from Wisconsin’s statewide cancer registry. Female controls of similar age and living in rural areas of Wisconsin were randomly selected from population lists (n=3,669). The addresses at diagnosis or reference year of study participants were assigned latitude/longitude coordinates (geocoded). The results from three statewide random studies of atrazine levels in well water in 1994, 1996, and 2001 were obtained from the Wisconsin Department of Agriculture, Trade, and Consumer Protection. Natural neighbor interpolation was used to estimate atrazine exposure levels separately for each of the 3 years. The mean atrazine exposure level was assigned to each participant based on her geocode. After adjustment for established breast cancer risk factors, compared to women in the lowest category of atrazine exposure (<0.15 ppb), the odds ratio of breast cancer for women exposed to atrazine concentrations of 1.0-2.9 ppb was 1.1 (95% CI 0.9-1.4). Results from this large population-based study do not suggest an increased risk of breast cancer from adult exposure to atrazine in drinking water. The possible risk for women exposed to levels of atrazine at or above statutory action levels of >or=3 ppb (OR 1.3, 95% CI 0.3-6.5) could not be ruled out due to small numbers in this category.