A feasibility study of phosphorylated-polyvinyl alcohol immobilized and free mixed bacterial culture bioaugmentation for removing atrazine in agricultural infiltrate was conducted utilizing a sand column setup. The effects of bacterial cell loading and infiltration rate on atrazine degradation were investigated by short-term tests in which the amount of synthetic infiltrate fed through was five times of the void volume (five pore volumes) of the sand column. In addition, the loss of the inoculated atrazine-degrading cultures and the change of bacterial community were determined. Selected tests were continued for monitoring a long-term performance of the system (50 pore volumes of the sand column). The results indicated that the inoculated cells removed 42-80% of the atrazine. The infiltration rate and cell loading significantly affected the atrazine removal. In the short-term tests, the immobilized and free cells provided similar atrazine removal; however, leaching of the free cells was Much greater than that of the immobilized cells. For the long-term performance. only the immobilized cells provided consistent atrazine removal efficiency throughout the test. Both immobilized and free cell systems exhibited a significant change in bacterial community structure during the atrazine degradation experiments. The infiltration rate was a significant factor for the change. (C) 2009 Elsevier B.V. All rights reserved.