The National Water-Quality Assessment (NAWQA) Program of the U.S. Geological Survey (USGS) developed an approach for a national assessment of the occurrence, status, and distribution of volatile organic compounds (VOCs) in samples of ground water from aquifer studies and in samples from drinking-water supply wells, specifically domestic and public wells. Data on VOCs in samples from aquifers and from domestic and public wells had three sources: (1) NAWQA sampling of ground water in aquifer studies and of shallow ground water in areas of agricultural or urban land use from 1993-2002, (2) retrospective data from other Federal, State, and local agencies that sampled ground water in aquifer studies from 1985-1997, and (3) sampling of ground water from public wells used as a source of drinking water from 1999-2000 (source-water survey). Data for assessing VOCs in ground water came from the NAWQA sampling of aquifers and from retrospective data (1 and 2). Domestic wells were the most commonly sampled well type in these aquifer studies. Data for assessing VOCs in domestic well samples came from the NAWQA sampling of aquifers and shallow ground water and from retrospective data (1 and 2). Data for assessing VOCs in public well samples came from NAWQA sampling of aquifers and shallow ground water and from the source-water survey (1, 2, and 3). Fifty-five VOCs were included in the national assessment. All ground-water samples were collected using consistent, prescribed field protocols and were analyzed using approved analytical methods. All ground-water collection activities and laboratory procedures included the collection of various quality-control samples in order to ensure the quality of the data. Comparisons of detection frequencies and detected concentrations among individual VOCs, groups of VOCs, aquifer studies, and data sets were performed using two assessment levels: (1) an assessment level of 0.2 microgram per liter (?g/L), and (2) an assessment level of 0.02 ?g/L. Selection of the 0.2-?g/L assessment level was based on historical laboratory reporting levels of VOCs used by the USGS National Water-Quality Laboratory (NWQL) prior to April 1996, and selection of the 0.02-?g/L assessment level was based on a new, low-level analytical method developed by the NWQL and used since April 1996. Relational analyses using multivariate logistic regression were performed on VOC data from aquifers, domestic wells, and public wells to better understand the natural and anthropogenic factors that control or influence the occurrence of VOCs. Ancillary data used in the relational analyses represent a variety of anthropogenic and hydrogeologic controls on the occurrence of VOCs. For aquifers, relational analyses were performed only on NAWQA data analyzed using the new low-level analytical method for 10 frequently detected VOCs. For domestic wells, relational analyses were performed at assessment levels of 0.2 ?g/L and 0.02 ?g/L for 6 and 10 frequently detected VOCs at these levels. For public wells, relational analyses were performed at an assessment level of 0.2 ?g/L for nine frequently detected VOCs.