Following the discovery of high concentrations of arsenic (up to 10,000 μg/liter) In the well water of a residential area near Fairbanks, Alaska, an epidemiologic study was undertaken in September, 1976, to assess exposure, absorption, and clinical sequelae of chronic arsenic ingestion. Two hundred eleven (91%) of 232 area residents completed questionnaires relating to water consumption history and to the signs and symptoms of arsenicalism. Physical examinations were conducted, and complete blood counts performed. Urine, hair, nail, and well water samples were analyzed for arsenic content. Urine arsenic levels above 2 μg/100 ml were found in 130/198 (66%) of the study population. Hair arsenic levels above 1 μg/g occurred in 74/181 (41%), and nail arsenic levels above 4 μg/g In 49/132 (37%). In well-water drinkers, a close correlation was found between well water arsenic levels and levels of arsenic In urine (r = 0.58, p < 10−8). Mean urine arsenic levels (17.83 μg/100 ml) In persons who drank well water containing ≥100μg arsenic/liter were significantly greater than levels in bottled-water drinkers or In well-water drinkers exposed to p < 10−10). Hair arsenic levels also correlated positively with well water arsenic levels (r = 0.43; p < 10−6), but may have reflected external contamination of hair during washing. Nail arsenic levels correlated poorly with well water arsenic exposure. Despite the chronic Increased exposure of the study population to arsenic, no clinical or hematologic abnormalities were found. Although the study did not consider long-term carcinogenic effects, based on the water level of arsenic (100 μg/liter) above which urine arsenic In drinkers Increases, the EPA standard (50 μg/liter) seems reasonable and conservative.