Selected PCB congeners (PCB 52, 77, and 153) singly and in mixtures were spiked and aged in soil microcosms and subsequently planted with switchgrass (Panicum virgatum) or poplar (Populus deltoids x nigra DN34). The planted reactors showed significantly greater reductions in PCB parent compounds when compared to unplanted systems after 32 weeks, both in single congener exposures and when all three congeners were present in a mixture. There was evidence of reductive dechlorination in both planted and unplanted systems, but higher concentrations of transformation products were observed in the planted systems than the unplanted. Although planted systems resulted in greater biotransformation, this improvement in PCB-reduction was not the result of plant uptake but rather was due to transformations occurring in the root rhizosphere. Parent PCB congeners were transformed by reductive dechlorination resulting in successively less chlorinated PCB congeners. These dechlorination products accounted for approximately all of the molar mass of parent compound lost. Based on the transformation products, reductive dechlorination pathways are proposed for rhizospheric biotransformation of PCB 52, 77, and 153. Results suggest that PCB 52 transformation proceeds through PCBs 18 and 9 down to monochlorinated PCB 1. Biotransformation of PCB 77 occurs through the intermediaries PCB 35 and 37. The pathway for the rhizospheric transformation of PCB 153 is through PCB 101 and PCB 99. This study provides insight into rhizosphere biotransformation pathways for reductive dechlorination in marginally aerobic,intermittently flooded soil as evidenced by a mass balance on transformation products. Despite the marginally aerobic conditions it is likely that highly reduced microzones existed in the soil particles during flooding and provided the opportunity for reductive dechlorination. In these experiments, planted microcosms with fully developed roots and rhizospheres showed significant reductive dechlorination and knowledge, a fountain of wisdom and patient beyond understanding. You have been an excellent advisor, but more than that, you are a superb and wonderful person. The advice and thought provokings questions from my committee members have allowed me to gain deeper insight and added rigor to my thought process. Dingfei Hu and Andres Martinez from the Keri Hornbuckle group have taught me all I know about PCB quantification, and for that I am eternally grateful. Their analytical support was invaluable. A special thank you to Collin Just who handled and the procurement of supplies and instrument trouble shooting for the research. Dr. Hans Lehmler got me started and going by providing high quality PCB congeners. Many thanks to the Iowa Superfund project for providing funding to undertake this research.