Soil contamination by organic compounds, especially by the petroleum-derived polycyclic aromatic hydrocarbons (PAHs), is a growing problem with serious environmental consequences. To evaluate the contribution of the soil macrofauna on the dissipation of PAHs in rhizosphere soil, a pot experiment was carried out to investigate effect of earthworm (Pheretima hupeiensis) on ryegrass (Lolium multiforum) remedying soils polluted by pyrene in, and roles of biotic & abiotic factors in the process of dissipation of pyrene in soils were estimated. After washing with sterile distilled water, eight adult earthworms (7 to 8 cm in length) were added on the top of sample soil of each treated pots at initial experiments,giving (a) unplanted microbe-inhibited pots with spiked soil (0.1% NaN₃ was used to inhibit the microbial activity), (b) unplanted pots with spiked soil, (c) planted microbe-inhibited pots with spiked soil (0.1% NaN₃ was used), (d) planted pots with spiked soil and (e) planted pots with unspiked soil, and each treatment was established with five replications. Results showed that earthworm activity promoted growth of ryegrass growing in soils at initial concentrations of pyrene ranging from 20.24 to 321.42 mg/kg, whose biomass per plant was 15.65%-21.32% larger than those in corresponding treatments without earthworm introduced in, and apparently facilitated its root/shoot ratio, which was 20.44% 24.24% greater than those in corresponding non-inoculated soils at the end of the experiment. Seventy-two days after earthworms were introduced in, residual pyrene concentrations (2.78 to 128.53 mg/kg) in vegetated soils were much lower than the respective initial values, and the values varied significantly among treatments. During the experiment, averagely 74.66% of pyrene was removed from the soils with plantation of grass and incubation of earthworm, which was 10.11% higher than those with only plantation of ryegrass, and 56.42% higher than those without plantation of ryegrass and earthworm inoculation. As compared to those treatments with same pollution levels, the contents of pyrene accumulated in plant tissues growing in soils with earthworms were always lower than those without earthworm. Along with the increments of soil pyrene concentrations from 20.24 to 321.42 mg/kg, the extractable pyrene in roots and shoots of ryegrass growing in soils without earthworms increased from 4.11 to 16.79 mg/kg and from 0.87 to 3.01 mg/kg, while they did from 2.32 to 11.17 mg/kg and from 0.51 to 2.01 mg/kg when earthworms were introduced in, respectively. Despite incubation of earthworm evidently enhanced the remediation of pyrene in soils, contributions of biotic and abiotic factors to phytoremediation process displayed distinct diversity. Of 10.11% enhanced removal of the total pyrene, abiotic loss, earthworm accumulation, phytodegredation and microbial degradation accounted for 0.16%, 0.08%, 0.42% and 2.64%, respectively. Among all the remedying factors, the plant-microbial interaction was proved to be the primary pathway for pyrene degradation, whether the presence of earthworms or not, which accounted for 51.75% of the total removal, and was 6.81% higher than those in corresponding soils without earthworm activity. These findings suggest a feasible way for the establishment of high-efficient phytoremediation of soil PAHs pollution by introducing earthworms into the soil-plant system.