Microcosms were used to investigate whether soil exposure to mulberry root extracts (rich in phenolic compounds) select for bacteria that degrade polynuclear aromatic hydrocarbons (PAHs). Unlike previous studies with freshly spiked soil, the present experiments were conducted with soils aged for 518 d with [14C]phenanthrene to decrease bioavailability and avoid exaggerating the selective pressure exerted by PAHs relative to the rhizosphere effect. Microcosms simulating contaminated planted soil were exposed to carbon at 20 mg/L/week of mulberry root extract for 211 d to simulate rhizodeposition. Contaminated bulk soils microcosms were amended with a C-free mineral medium to discern the effect of rhizodeposition. Uncontaminated soil controls also were exposed to similar dose regimes. Real-time quantitative polymerase chain reaction was used to enumerate total bacteria and PAH degraders harboring the genes nahAc (coding for naphthalene dioxygenase), todC1 (coding for toluene/benzene/chloro-benzene dioxygenase), bmoA (coding for hydroxylating monooxygenases), and dmpN (coding for phenol hydroxylase). Exposure to root extracts enhanced the growth of total bacteria and PAH degraders in both contaminated and uncontaminated rhizosphere microcosms. The relative abundance of PAH-degrader gene copies (as a fraction of the total bacteria) was similar for different treatments, suggesting that the root extracts did not select for PAH degraders. Overall, these results suggest that rhizodeposition from phenolic releasers contributes to the fortuitous (but not selective) proliferation of PAH degraders, which may enhance phytoremediation.