The potential for herbicide degradation in an unconfined chalk aquifer was examined by collecting and spiking fresh samples and incubating them in the laboratory. The microcosms were incubated at 20 degrees C under aerobic conditions and spiked with either isoproturon, atrazine or mecoprop at a concentration of 100 mu g/l. The samples were obtained from a single fieldsite within the Upper Chalk aquifer in Hampshire, UK. Groundwater samples required the presence of sterile chalk in a ratio of at least 1:13 to promote isoproturon degradation. An isoproturon degradation potential existed in the soil, and the chalk unsaturated and saturated zones. However, no degradation of isoproturon in the unsaturated zone was observed when a more appropriate simulation of in-situ moisture conditions was carried out. Apart from the soil, no potential for atrazine or mecoprop degradation could be detected in the same samples over a 200-day incubation. In a series of groundwater samples taken from different boreholes, 10-300 m apart, large differences in isoproturon degradation potential were observed. Removal rates for 100 mu g/l isoproturon varied from 83-425 ng/day, but in some samples no degradation potential could be detected. The primary metabolite which could be distinguished from isoproturon degradation in chalk and groundwater was monodesmethyl-isoproturon. When a chalk groundwater sample was spiked with isoproturon at 0.9 mu g/l, this was not degraded over a 300-day incubation period. Further experiments with fresh groundwater from a Triassic Sandstone site illustrated that groundwater bacteria could degrade isoproturon at the more realistic temperature of 10 degrees C as well as at 20 degrees C. (C) 2000 Elsevier Science B.V. All rights reserved.