Strychnine (Strychnidin-10-one) was used as a rodenticide to control a mouse plague across southern Australia in 1993. Wheat grains were coated with a mixture (containing strychnine, sugar, flour, oil, and a dye) at a rate of 3 g strychnine/kg grain, and were spread as a bait for mice at a density of approximately 3 treated grains/m2. We examined the release of strychnine residue from the treated grains into soils and its sorption-desorption behaviour in 4 different South Australian soils, ranging from sand to clay in texture and acidic to alkaline in reaction. Release studies using treated grains showed that, unless the treated grains were buried in soil, the release of strychnine from the treated grain into soil was slow, requiring about 2 months for its complete transfer into soil. However, upon burial of the treated grain, such as during resowing, the transfer of strychnine from the treated grain to soil was rapid (>90% within 7 days). The sorption of strychnine was found to occur rapidly (>90% of sorption within 15 min). The sorption affinity of strychnine varied by almost 1 order of magnitude among the 4 soils studied. An acidic clay soil (Mintaro) sorbed nearly all (97%) of the applied strychnine, whereas the alkaline sandy soil (Bute) sorbed only 54% of the applied amount, when a strychnine solution 10 mg/L was equilibrated with the soil. Sorption was found to increase substantially with decreasing pH(Ca) from 9 to 7. Indeed, in the 2 soils with relatively higher sorption capacities, nearly 100% of the applied strychnine was sorbed at pH(Ca) 6.5. The pH dependency of sorption was found to follow the proportion of cationic species of strychnine base present in the soil solution. During desorption, hysteresis between the sorption and desorption isotherms of strychnine was noted. The study showed that except in very sandy, alkaline soils, the sorption of strychnine is likely to be high.