The degradation of strychnine a rodenticide used to control a major mouse plague in South Australia during 1993 was studied under laboratory conditions in three agricultural soils (Bute, Booleroo and Mintaro) with contrasting physico-chemical properties (pH, mineralogy, organic matter). Strychnine disappeared rapidly in non-sterile (biologically active) Bute and Booleroo soils, from an initial concentration of 50 mg kg-1 soil to less than 1 mg kg-1 within 42 d. Strychnine was not degraded in sterilized Booleroo soil, suggesting that microbiological processes were responsible for the degradation of strychnine in these two alkaline soils. Degradation of strychnine in both soils was modelled using a logistic regression model, and was characterized by an 8-14 d 'lag phase' followed by rapid strychnine disappearance. In contrast strychnine degradation was not observed both in the non-sterile and sterilized Mintaro soil during a 98 d incubation. The different pattern of strychnine degradation between these soils is discussed in terms of the sorption behaviour and bio-availability of strychnine. It is suggested that in the acid pH Mintaro soil (pH 5.5 ([10 mM CaCl2])), degradation is unable to proceed most likely due to low strychnine bio-availability, as a result of strong strychnine adsorption on to the soil organic and mineral phases.