As part of the U.S. Geological Survey's Resource Conservation and Recovery Act, Facilities Investigations at Fort Bragg, North Carolina, selected geophysical techniques were evaluated for their usefulness as assessment tools for determining subsurface geology, delineating the areal extent of potentially contaminated landfill sites, and locating buried objects and debris of potential environmental concern. Two shallow seismic-reflection techniques (compression and shear wave) and two electromagnetic techniques (ground-penetrating radar and terrain conductivity) were evaluated at several sites at the U.S. Army Base. The electromagnetic techniques also were tested for tolerance to cultural noise, such as nearby fences, vehicles, and power lines. For the terrain conductivity tests, two instruments were used--the EM31 and EM34, which have variable depths of exploration. The shallowest reflection event was 70 feet below land surface observed in common-depth point, stacked compression-wave data from 24- and 12-fold shallow-seismic-reflection surveys. Several reflection events consistent with clay-sand interfaces between 70 and 120 feet below land surface, along with basement-saprolite surfaces, were imaged in the 24-fold, common- depth-point stacked data. 12-fold, common-depth-point stacked data set contained considerably more noise than the 24-fold, common-depth-point data, due to reduced shot-to-receiver redundancy. Coherent stacked reflection events were not observed in the 24-fold, common-depth-point stacked shear-wave data because of the partial decoupling of the shear- wave generator from the ground. At one site, ground-penetrating radar effectively delineated a shallow, 2- to 5-foot thick sand unit bounded by thin (less than 1 foot) clay layers. The radar signal was completely attenuated where the overlying and underlying clay units thickened and the sand unit thinned. The pene- tration depth of the radar signal was less than 10 feet below land surface. A slight increase in electromagnetic conductivity across shallow sampling EM31 and EM34 profiles provided corroborative evidence of the shallow, thickening clay units. Plots of raw EM31 and EM34 data provided no direct interpretable information to delineate sand and clay units in the shallow subsurface. At two sites, the ground-penetrating radar effectively delineated the lateral continuity of surficial sand units 5 to 25 feet in thickness and the tops of their underlying clay units. The effective exploration depth of the ground-penetrating radar was limited by the proximity of clay units to the subsurface and their thickness. The ground-penetrating radar delineated the areal extent and depth of cover at a previously unrecognized extension of a trench-like landfill underlying a vehicle salvage yard. Attenuation of the radar signal beneath the landfill cover and the adjacent subsurface clays made these two mediums indistinguishable by ground-penetrating radar; however, EM31 data indicated that the electrical conductivity of the landfill was higher than the subsurface material adjacent to the landfill. The EM31 and EM34 conductivity surveys defined the areal extent of a landfill whose boundaries were inaccurately mapped, and also identified the locations of an old dumpsite and waste incinerator site at another landfill. A follow-up ground-penetrating radar survey of the abandoned dumpsite showed incongruities in some of the shallow radar reflections interpreted as buried refuse dispersed throughout the landfill. The ground-penetrating radar and EM31 effectively delineated a shallow buried fuel-oil tank. Of the three electromagnetic instruments, the ground-penetrating radar with the shielded 100-megahertz antenna was the least affected by cultural noise followed, in order, by the EM31 and EM34. The combination of terrain- conductivity and ground-penetrating radar for the site assessment of the landfill provided a powerful means to identify the areal extent of the landfill, potenti