We report the first data for atrazine removal in low-turbidity freshwaters. Atrazine is a globally applied herbicide, contamination by which may lead to direct and indirect ecotoxicological impacts. Although a common contaminant of surface waters, microbial biodegradation of atrazine in this environment has been little studied, with most work focused on soils by means of selected, atrazinedegrading bacteria-enriched cultures. Here, we measured atrazine removal from river water using a batch incubation system designed to represent environmental conditions, with water from two contrasting UK rivers, the Tamar and Mersey. Atrazine and bacterial inocula prepared from the source water were added to cleaned river water for 21-day incubations that were analysed directly by electrospray ionisation-mass spectrometry. The experimental approach was validated using peptides of different molecular mass. Results show that atrazine concentrations decreased by 11% over 21 days in Tamar samples, a rural catchment with low population density, when atrazine was the only substrate added. In contrast no removal was evident in Mersey samples, an urban catchment with high population density. When a tripeptide was added as a co-substrate, atrazine removal in the Tamar water remained at 11% while that for the Mersey water increased from 0 to 37%. Although degradation of atrazine in aerobic freshwaters is predicted according to its chemical structure, our data suggest that the composition of the bacterial population determines whether removal occurs under these conditions and at these environmentally realistic concentrations.