Female houseflies (Musca domestica L.) from a susceptible and a multi-insecticide-resistant strain were used to evaluate the relative toxicity of an insecticide bait formulation of the carbamate insecticide methomyl. Individual flies were allowed to feed on bait granules for an unrestricted period or for 5 s. Resistant flies took longer than susceptible flies to initiate a feeding response. When allowed to feed continuously, those from the resistant strain spent longer feeding than susceptible ones. The time taken to knock-down (KD), including feeding times, was significantly greater for resistant than susceptible flies (P less than 0.001), but once the proboscis was withdrawn from the granule there was no difference in KD times between the strains. All flies from both strains were knocked down, and only a very small number of resistant and susceptible flies recovered. The toxic effects of methomyl on flies which were restricted to a 5 s feed ranged from no observed effect to KD in less than 1 min. After feeding for 5 s, 81% of resistant and 98% of susceptible flies developed signs of methomyl poisoning. More resistant than susceptible flies recovered from KD, giving final mortalities of 46% and 88% respectively. With both feeding regimes, some flies of both strains which had apparently recovered from KD had lost their ability to fly. Observations have also shown that 8% of resistant flies may have been repelled by methomyl granules. The implication of these results on the survival of M. domestica in intensive animal units following exposure to methomyl bait is also discussed.