A regional study of the areal and depth distribution of pesticides in shallow groundwater in the Delmarva Peninsula of Delaware, Maryland, and Virginia was done to (i) relate the pesticides detected to landscape and shallow subsurface features, and (ii) evaluate aquifer vulnerability and the potential contamination of drinking-water supplies. Water samples collected at 100 wells from 1988 to 1990 were analyzed for concentrations of 36 pesticides, four metabolites, and other constituents. The most commonly detected residues were atrazine, cyanazine, simazine, alachlor, metolachlor, and dicamba. Concentrations were low; few exceeded 3 µg L⁻¹. Most detections correlate with the intensive use of these herbicides in three widely distributed and commonly rotated crops—corn (Zea mays L.), soybean [Glycine max (L.) Merr.], and small grain—particularly if grown in well-drained soils. Most detections occurred in samples collected from shallow wells screened within 10 m of the overlying water table. The shallow depth distribution of most residues is consistent with their suspected history of use (ca. 20 yr), and patterns in shallow groundwater flow in the surficial aquifer in the study area. The areal and depth distributions of detectable residues in groundwater did not correlate with a vulnerability index, nor any of the component scores developed to estimate that index using the DRASTIC method. The shallow depth of most detections also indicates why few samples from water-supply wells in this study had measurable concentrations of pesticides; most supply wells are deeper than 10 m below the water table. The low number of contaminated samples from supply wells implies that deep groundwater currently (1992) used for drinking generally does not contain detectable pesticide residues.