none Laboratory bioassays were carried out to determine the toxicity of residues of the synthetic pyrethroid insecticide, deltamethrin ((S)- alpha -cyano-3-phenoxybenzyl (1R)-cis-3-(2,2-dibromovinyl)-2,2-dimethylcyclopropanecarboxylate) to a range of beneficial invertebrates exposed to treated wheat foliage, sandy loam soil, and glass surfaces. The test invertebrates represented a range of predators and a parasitoid that inhabit the foliage and grounds layers of temperate cereal crops, i.e., Pterostichus melanarius (Illiger), Nebria brevicollis (F.), Demetrias atricapillus (L.) and Bembidion obtusum (Serville) (Coleoptera: Carabidae), Tachyporus hypnorum (F.) (Coleoptera: Staphylinidae), Coccinella septempunctata (L.) (Coleoptera: Coccinellidae) adults (A) and larvae (L), Episyrphus balteatus (Degeer) (Diptera: Syrphidae), and Aphidius rhopalosiphi (De Stefani-Perez) (Hymenoptera: Braconidae). In 2 h flag leaf exposure bioassays, the LD sub(50) values varied from 0.4 g ai/ha to >50 g ai/ha. The susceptibility ranking, from most to least susceptible, was C. septempunctata (L) > T. hypnorum > C. septempunctata (A) > E. balteatus > A. rhopalosiphi > D. atricapillus. On soil, the LD sub(50) values were 52.8 g ai/ha for T. hypnorum and 97.8 g ai/ha for C. septempunctata (A) after 2 h exposure. This period of exposure was, however, found to be of insufficient duration to separate the susceptibilities of the other test species adequately, and therefore 72 h exposure bioassays were also carried out. The 72 h LD sub(50) values varied between 4.2 g ai/ha and 267 g ai/ha. The susceptibility ranking, from most to least susceptible, was T. hypnorum > B. obtusum > C. septempunctata (A) > P. melanarius > N. brevicollis > D. atricapillus. In 2 h glass bioassays the LD sub(50) range was 1.2 g ai/ha to >37 g ai/ha and the susceptibility ranking was T. hypnorum > C. septempunctata (A) > N. brevicollis, and D. atricapillus. Residual toxicities in the different 2 h exposure bioassays were compared for T. hypnorum and C. septempunctata (A) by iterating sequences of lethal dose ratios between LD sub(10) and LD sub(90) from the dose-response statistics for each respective pair of substrates.