We investigated the effects of vegetation on the fate of pentachlorophenol (PCP) in soil using a novel high-flow sealed test system. Pentachlorophenol has been widely used as a wood preservative, and this highly toxic biocide contaminates soil and ground water at many sites. Although plants are known to accelerate the rates of degradation of certain soil contaminants, this approach has not been thoroughly investigated for PCP. The fate of [14C]PCP, added to soil at a concentration of 100 mg/kg, was compared in three unplanted and three planted systems. The plant used was Hycrest, a perennial, drought-tolerant cultivar of crested wheatgrass [Agropyron desertorum (Fischer ex Link) Schultes]. The flow-through test system allowed us to maintain a budget for 14C-label as well as monitor mineralization (breakdown to 14CO2) and volatilization of the test compound in a 155-d trial. In the unplanted systems, an average of 88% of the total radiolabel remained in the soil and leachate and only 6% was mineralized. In the planted system, 33% of the radiolabel remained in the soil plus leachate, 22% was mineralized, and 36% was associated with plant tissue (21% with the root fraction and 15% with shoots). Mineralization rates were 23.1 mg PCP mineralized kg-1 soil in 20 wk in the planted system, and for the unplanted system 6.6 mg PCP kg-1 soil for the same time period. Similar amounts of volatile organic material were generated in the two systems (1.5%). Results indicated that establishing crested wheatgrass on PCP-contaminated surface soils may accelerate the removal of the contaminant.