A source- and finished-water-quality assessment of groundwater was conducted in the Piedmont Physiographic Province of Maryland and Virginia in the Potomac River Basin during 2003-04 as part of the U.S. Geological Survey's National Water-Quality Assessment Program. This assessment used a two-phased approach to sampling that allowed investigators to evaluate the occurrence of more than 280 anthropogenic organic compounds (volatile organic compounds, pesticides and pesticide degradates, and other anthropogenic organic compounds). Analysis of waters from 15 of the largest community water systems in the study area were included in the assessment. Source-water samples (raw-water samples collected prior to treatment) were collected at the well head. Finished-water samples (raw water that had been treated and disinfected) were collected after treatment and prior to distribution. Phase one samples, collected in August and September 2003, focused on source water. Phase two analyzed both source and finished water, and samples were collected in August and October of 2004. The results from phase one showed that samples collected from the source water for 15 community water systems contained 92 anthropogenic organic compounds (41 volatile organic compounds, 37 pesticides and pesticide degradates, and 14 other anthropogenic organic compounds). The 5 most frequently occurring anthropogenic organic compounds were detected in 11 of the 15 source-water samples. Deethylatrazine, a degradate of atrazine, was present in all 15 samples and metolachlor ethanesulfonic acid, a degradate of metolachlor, and chloroform were present in 13 samples. Atrazine and metolachlor were present in 12 and 11 samples, respectively. All samples contained a mixture of compounds with an average of about 14 compounds per sample. Phase two sampling focused on 10 of the 15 community water systems that were selected for resampling on the basis of occurrence of anthropogenic organic compounds detected most frequently during the first phase. A total of 48 different anthropogenic organic compounds were detected in samples collected from source and finished water. There were a similar number of compounds detected in finished water (41) and in source water (39). The most commonly detected group of anthropogenic organic compounds in finished water was trihalomethanes - compounds associated with the disinfection of drinking water. This group of compounds accounted for 30 percent of the detections in source water and 44 percent of the detections in finished water, and were generally found in higher concentrations in finished water. Excluding trihalomethanes, the number of total detections was about the same in source-water samples (33) as it was in finished-water samples (35). During both phases of the study, two measurements for human-health assessment were used. The first, the Maximum Contaminant Level for drinking water, is set by the U.S. Environmental Protection Agency and represents a legally enforceable maximum concentration of a contaminant permitted in drinking water. The second, the Health-Based Screening Level, was developed by the U.S. Geological Survey, is not legally enforceable, and represents a limit for more chronic exposures. Maximum concentrations for each detected compound were compared with either the Maximum Contaminant Level or the Health-Based Screening Level when available. More than half of the compounds detected had either a Maximum Contaminant Level or a Health-Based Screening Level. A benchmark quotient was set at 10 percent (greater than or equal to 0.1) of the ratio of the detected concentration of a particular compound to its Maximum Contaminant Level, or Health-Based Screening Level. This was considered a threshold for further monitoring. During phase one, when only source water was sampled, seven compounds (chloroform, benzene, acrylonitrile, methylene chloride, atrazine, alachlor, and dieldrin) met or exceeded a benchmark quotient.