Purpose The assessing bias of rhizosphere effect on polycyclic aromatic hydrocarbons (PAHs) degradation in soils would come out from formation of nonextractable PAHs and extractability difference of various solvents. The aim of this study was to evaluate the role of rhizosphere effect in longterm PAHs polluted soils by using sequential extraction approach. Material and methods The scheme of sequential extraction included methanol/water extractable PAHs, butanol extractable PAHs, DCM extractable PAHs, humic acid-bound PAHs, crude humin-bound PAHs, and organic-C enriched humin-bound PAHs. PAHs in plant tissues were extracted by dichloromethane after saponifying. The correlations between PAHs in plant tissues and sequentially extracted fractions were generated by partial least squares regression. Results and discussion The profiles of sequentially extracted PAHs varied with plant species. The discrepancy of toxicity equivalency concentrations between rhizosphere and bulk soils was much more significant than that of total PAHs concentrations. In partial least squares regression models, the concentration of PAHs in plant tissues was correlated with fractions strongly associated with soil.Conclusions The novelty of this study is the evaluation of concentration and toxicity equivalency concentration of PAHs in rhizosphere of crops sampled in a field polluted with PAHs for long term. This study has highlighted more significant role of rhizosphere in cleanup of cancerogenic toxicity of soil than amount of PAHs in polluted soils.