Malaria vector mosquitoes belonging to the Anopheles gambiae complex were studied in four hamlets in The Gambia. All inhabitants were given bednets treated either with a placebo (milk) in two hamlets or with the pyrethroid insecticide permethrin (500 mg/m2) in two other hamlets. Malaria transmission occurred mainly during a few weeks of the rainy season, in September and October 1987. The indoor resting densities of mosquitoes in permethrin-treated hamlets were reduced, and we estimated over 90% reduction in biting on man by An. gambiae Giles sensu stricto in these hamlets. No mosquitoes were found under permethrin-treated bednets compared with eighty-one recovered from placebo-treated bednets. Mosquitoes exited more readily from rooms where permethrin-treated bednets were used than from rooms with placebo-treated nets. The annual mean probability that a child would receive an infective bite was estimated to be 0.09 in hamlets with insecticide-treated bednets, compared with 1.9 where placebo-treated bednets were used. Permethrin-treated bednets are therefore recommended as a means of effectively reducing the risk of exposure to malaria transmission, particularly in areas of low seasonal transmission.