In laboratory incubation experiments, application of a commercial formulation of the herbicide butachlor (N-butoxymethyl-2-chloro-2',6'-diethyl acetanilide) to three tropical rice soils, widely differing in their physicochemical characteristics, under flooded condition inhibited methane (CH4) production. The inhibitory effect was concentration dependent and most remarkable in the alluvial soil. Thus, following application of butachlor at 5, 10, 50 and 100 microg g(-1) soil, respectively, cumulative CH4 production in the alluvial soil was inhibited by 15%, 31%, 91% and 98% over unamended control. Since CH4 production was less pronounced in the sandy loam and acid sulfate soil, the impact of amendment with butchalor, albeit inhibitory, was less extensive than the alluvial soil. Inhibition of CH4 production in butachlor-amended alluvial soil was related to the prevention in the drop in redox potential as well as low methanogenic bacterial population especially at high concentrations of butachlor. CH4 oxidation was also inhibited in butachlor-amended alluvial soil with the inhibitory effect being more prevalent under flooded condition. Inhibition in CH4 oxidation was related to a reduction in the population of soluble methane monooxygenase producing methanotrophs. Results demonstrate that butachlor, a commonly used herbicide in rice cultivation, even at very low concentrations can affect CH4 production and its oxidation, thereby influencing the biogeochemical cycle of CH4 in flooded rice soils.