Studies of disinfection byproducts ( DBPs ) in drinking water and risk of adverse reproductive outcome have usually relied on approximate measures of exposure. Individual differences in consumption of bottled or filtered water, variability in tap water consumption at home and at work, dermal and inhalation exposure to volatile contaminants, and changes in residency during pregnancy may lead to exposure misclassification. We characterized exposures to tap water and other risk factors among 71 pregnant and 43 non -pregnant women attending public health clinics. Nearly all residences had a municipal water source, but 25% of women drank filtered or bottled water. Fifty percent of the women in our sample reported working outside the home where, on average, one third of their daily water intake took place. Pregnant women consumed more water than non -pregnant women ( 3.4 vs. 3.0 total l / day ) , especially cold tap water at home ( 1.8 vs. 1.3 l / day, 95% CI for the difference = 0.1, 0.9 ) . Patterns of showering were similar for both groups of women, but pregnant women were more likely to bathe and to bathe more frequently. The prevalence of smoking was lower among pregnant women ( 22.5% vs. 32.6% ) , as was the consumption of alcohol ( 4.2% vs. 53.5%, 95% CI for the difference = Ă 64.9, Ă 33.7 ) . Thirty -two percent of women had moved during their current pregnancy. The data reaffirm the importance of collecting individual -level data for water consumption and exposure to potential confounders to avoid misclassification bias. This study is the first to target women of low socio -economic status ( SES ) and therefore of particular interest in studies of adverse reproductive outcomes for which this group is at increased risk.