Test methods using common laboratory-reared, nonsoil insects such as southern armyworms, Prodenia eridania (Cramer) ; house fly larvae, Musca domestica L.; and confused flour beetles, Tribolium confusum Jacquelin duVal, to indicate effective soil insecticides are described. Using these new test methods and other evaluations, Dursban® insecticide (O,O-diethyl O-3,5,6-trichloro-2-pyridyl phosphorothioate) was found to be superior to 31 other commercial insecticides (24 organophosphorus insecticides, 3 carbamates, and 4 chlorinated hydrocarbons). The activity of Dursban was surpassed only by isobenzan, a chlorinated hydrocarbon insecticide. Insecticidal performance of Dursban was unaffected by microorganisms, fertilizers, and growing plants. Residual activity, which ranged from 9 weeks at 1 lb per acre to 1 year at 20 lb per acre, was shortened a little by moisture and higher temperatures. Dursban apparently is bound very tightly to particles of soil and organic matter as evidenced by only slight volatilization from soil and its resistance to leaching from soil by water. Initial and residual activities of Dursban were modifiable by changes in formulation.