The ecological impact offour insecticides used to control desert locustswas investigated during experimental field trials in natural temporary ponds in a cultivated savannah area of Senegal, West Africa. The insecticide formulations, dose rates, and application techniques were similar to those used in operational desert locust control. Average initial concentrations of fenitrothion, diflubenzuron, deltamethrin, and bendiocarb in pond water were 80, 10.4, 0.45, and 24.0 μg/L, respectively. Diflubenzuron and deltamethrin virtually disappeared in 24 h. Pseudo–first-order half-lives were 34 h for fenitrothion and 17 d for bendiocarb. Fenitrothion and deltamethrin significantly reduced population densities of backswimmers of the genus Anisops (Hemiptera, No-tonectidae) and in addition caused an extensive kill of other species of insects. Both insecticide applications were also followed by reductions of zooplankton densities, especially Cladocera, and deltamethrin eradicated populations of fairy shrimp (Streptoce-phalus spp.: Branchiopoda, Anostraca). Diflubenzuron only affected crustaceans, i.e., cladocerans and fairy shrimp. The least harmful insecticide in the study was bendiocarb, which only reduced the density of cladocerans. Recovery proceeded at fixed rates, which were different for each taxon. Anisops spp. recovered from the treatments in 1 to 4 weeks, most likely through aerial migration. Cladocerans returned to normal densities in 3 to 6 weeks. Streptocephalus spp., however, whose resting eggs presumably need desiccation during the dry season to be able to hatch, did not reappear until the next rainy season, one year later. It is concluded that contamination of temporary ponds in the Sahel by insecticide drift from desert locust control operations with any of the four insecticides should be avoided.