The relationship between community drinking water quality and the occurrence of late adverse pregnancy outcomes was investigated by conducting a case-control study among women who delivered infants during August 1977 through March 1980 at Brigham and Women's Hospital in Massachusetts. The water quality indices were compared among 1,039 congenital anomaly cases, 77 stillbirth cases, 55 neonatal death cases, and 1,177 controls. Trace element levels were gathered from routine analyses of public water supplies from the communities in which the women resided during pregnancy. It was observed that, after adjustment for confounding, the frequency of stillbirths was increased for women exposed to chlorinated surface water (OR 2.6 95% CI 0.9-7.5) and for women exposed to detectable lead levels (OR 2.1; 95% CI 0.6-7.2); the frequency of cardiovascular defects was increased relative to detectable lead levels (OR 2.2, 95% CI 0.9-5.7); and the frequency of central nervous system defects was increased relative to the highest tertile of potassium (OR 6.3, 95% CI 1.1-37.3). The frequency of ear, face, and neck anomalies was increased in relation to detectable silver levels (OR 3.3, 95% CI 0.9-12.2), but the frequency decreased relative to high potassium levels (OR 0.2, 95% CI 0.1-0.7). The frequency of neonatal deaths was decreased relative to detectable fluoride levels (OR 0.4, 95% CI 0.2-1.0), and the frequency of musculoskeletal defects was decreased relative to detectable chromium levels (OR 0.4, 95% CI 0.2-1.0). The majority of these associations were not stable statistically. Further research is needed to corroborate these findings.