The degradation rate of atrazine in floodplain soils under natural grasslands and cropped fields in the Liverpool Plains, NSW, was studied under laboratory incubation and in glasshouse bioassays, and related to soil properties including microbial community analysis by t-RFLP. The experiments were part of a broader study aiming to manage pesticides in the environment using vegetative filters (biofilters). The soils differed in their atrazine treatment history. Degradation rate (half-life) in cropped soil was more rapid (≈2 to 7 days) than in 2 grassland soils (≈8 to ≈22 days). Bioassays in the glasshouse using oats and soybeans supported this finding. The t-RFLP analysis disclosed the presence of 2 consortia of bacterial species that are reported in the literature to degrade atrazine. These were: (i) Rhodococcus sp, Pseudomonas aeruginosa, and Clavibacter michiganense and (ii) Acinetobacter sp., Pseudomonas sp., and Streptomyces sp. Their dynamics during incubation suggested that they might have been responsible for the more rapid atrazine degradation in the cropped soil. The enhanced atrazine degradation in cropped soil was also associated with lower concentrations of soil organic C and percentage of light fraction carbon compared with grassland soils, suggesting that atrazine provided an additional source of N and C to organisms that can quickly degrade the herbicide. The finding of relatively short atrazine half-life has implications for the effectiveness of the herbicide, as well as for the loads of pesticide potentially entering the environment. The results suggest there is little risk of atrazine accumulating in biofilters and causing damage.