Rhizodegradation is a promising cleanup technology where microorganisms degrade soil contaminants in the rhizosphere. A symbiotic relationship is expected to occur between plant roots and soil microorganisms in contaminated soils that enhances natural microbial degradation. However, little is known about how different initial microbiotas influence the rhizodegradation outcome. Recent studies have hinted that soil initial diversity has a determining effect on the outcome of contaminant degradation. To test this, we either planted (P) or not (NP) balsam poplars () in two soils of contrasting diversity (agricultural and forest) that were contaminated or not with 50 mg kg of phenanthrene (PHE). The DNA from the rhizosphere of the P and the bulk soil of the NP pots was extracted and the bacterial genes encoding the 16S rRNA, the PAH ring-hydroxylating dioxygenase alpha subunits (PAH-RHDα) of Gram-positive and Gram-negative bacteria, and the fungal ITS region were sequenced to characterize the microbial communities. The abundances of the PAH-RHDα genes were quantified by real-time quantitative PCR. Plant presence had a significant effect on PHE degradation only in the forest soil, whereas both NP and P agricultural soils degraded the same amount of PHE. Fungal communities were mainly affected by plant presence, whereas bacterial communities were principally affected by the soil type, and upon contamination the dominant PAH-degrading community was similarly constrained by soil type. Our results highlight the crucial importance of soil microbial and physicochemical characteristics in the outcome of rhizoremediation. Polycyclic aromatic hydrocarbons (PAH) are a group of organic contaminants that pose a risk to ecosystems' health. Phytoremediation is a promising biotechnology with the potential to restore PAH-contaminated soils. However, some limitations prevent it from becoming the remediation technology of reference, despite being environmentally friendlier than mainstream physicochemical alternatives. Recent reports suggest that the original soil microbial diversity is the key to harnessing the potential of phytoremediation. Therefore, this study focused on determining the effect of two different soil types in the fate of phenanthrene (a polycyclic aromatic hydrocarbon) under balsam poplar remediation. Poplar increased the degradation of phenanthrene in forest, but not in agricultural soil. The fungi were affected by poplars, whereas total bacteria and specific PAH-degrading bacteria were constrained by soil type, leading to different degradation patterns between soils. These results highlight the importance of performing preliminary microbiological studies of contaminated soils to determine whether plant presence could improve remediation rates or not.