BACKGROUND: It is anticipated that malaria elimination efforts in Africa will be hampered by increasing resistance to the limited arsenal of insecticides approved for use in public health. However, insecticide susceptibility status of vector populations evaluated under standard insectary test conditions can give a false picture of the threat, as the thermal environment in which the insect and insecticide interact plays a significant role in insecticide toxicity. METHODS: The effect of temperature on the expression of the standard WHO insecticide resistance phenotype was examined using Anopheles arabiensis and Anopheles funestus strains: a susceptible strain and the derived resistant strain, selected in the laboratory for resistance to DDT or pyrethroids. The susceptibility of mosquitoes to the pyrethroid deltamethrin or the carbamate bendiocarb was assessed at 18, 25 or 30 degrees C. The ability of the pyrethroid synergist piperonyl-butoxide (PBO) to restore pyrethroid susceptibility was also assessed at these temperatures. RESULTS: Temperature impacted the toxicity of deltamethrin and bendiocarb. Although the resistant An. funestus strain was uniformly resistant to deltamethrin across temperatures, increasing temperature increased the resistance of the susceptible An. arabiensis strain. Against susceptible An. funestus and resistant An. arabiensis females, deltamethrin exposure at temperatures both lower and higher than standard insectary conditions increased mortality. PBO exposure completely restored deltamethrin susceptibility at all temperatures. Bendiocarb displayed a consistently positive temperature coefficient against both susceptible and resistant An. funestus strains, with survival increasing as temperature increased. CONCLUSIONS: Environmental temperature has a marked effect on the efficacy of insecticides used in public health against important African malaria vectors. Caution must be exercised when drawing conclusions about a chemical's efficacy from laboratory assays performed at only one temperature, as phenotypic resistance can vary significantly even over a temperature range that could be experienced by mosquitoes in the field during a single day. Similarly, it might be inappropriate to assume equal efficacy of a control tool over a geographic area where local conditions vary drastically. Additional studies into the effects of temperature on the efficacy of insecticide-based interventions under field conditions are warranted.