To transition catalytic reductive dechlorination (CRD) into practice, it is necessary to demonstrate the effectiveness, robustness, and economic competitiveness of CRD-based treatment systems. A CRD system scaled up from previous laboratory studies was tested for remediating groundwater contaminated with 500-1200 microg L(-1) trichloroethylene (TCE) at Edwards Air Force Base (AFB), California. Groundwater was pumped from a treatment well at 2 gal min(-1), amended with hydrogen to 0.35 mg L(-1) and contacted for 2.3 min with 20 kg eggshell-coated Pd on alumina beads (2% Pd by wt) packed in a fixed-bed reactor, and then returned to the aquifer. Operation was continuous for 23 h followed a 1 h regeneration cycle. After regeneration, TCE removal was 99.8% for 4 to 9 h and then declined to 98.3% due to catalyst deactivation. The observed catalyst deactivation was tentatively attributed to formation of sulfidic compounds; modeling of catalyst deactivation kinetics suggests the presence of sulfidic species equivalent to 2-4 mg L(-1) hydrogen sulfide in the reactor water. Over the more than 100 day demonstration period, TCE concentrations in the treated groundwater were reduced by >99% to an average concentration of 4.1 microg L(-1). The results demonstrate CRD as a viable treatment alternative technically and economically competitive with activated carbon adsorption and other conventional physicochemical treatmenttechnologies.