In a biopurification system such as a biobed, the rhizosphere of the grass layer may be a significant factor for promoting pesticide dissipation in the biomixture. The rhizosphere effect of a Lolium perenne, Festuca arundinacea and Trifolium repens mixture on the dissipation of a pesticide combination that was composed of atrazine, chlorpyrifos and isoproturon was studied. The assay was performed using glass pots divided into two separate compartments (root surface and root-free), each filled with an organic biomixture (oat husk, top soil and peat) and contaminated with the pesticide mixture at 5 mg kg(-1). Non-planted and non-contaminated pots were also used as controls. The results indicated that there were high atrazine, chlorpyrifos and isoproturon dissipation in the planted pots compared with the unplanted pots. An inverse correlation was found throughout the assay between phenoloxidase activity and residual pesticide (r=0.684 to 0.952). Indeed, fungal biomass was positively correlated with phenoloxidase activity on day 1 (r =0.825) and day 30 (r =0.855). Besides, exudation of oxalic and malic acid in contaminated pots was higher than in the control without pesticides, associated with oxidation of the pesticide mixture in the biomixture of a bioded system. Therefore, the grass layer enhances pesticide removal in biobeds.