Abstract.A village‐scale field trial of pyrethroid‐impregnated mosquito nets was undertaken in The Gambia, West Africa, in the Mandinka village of Saruja (13o13'N, 14o55'W) during July‐November 1989. Nearly all the villagers possessed and used their own bednets.Anopheles gambiaeis the main vector of human malaria in the area.An experimental wash‐resistant formulation of permethrin was compared with standard emulsifiable concentrate (EC) formulations of permethrin and lambda‐cyhalothrin, versus placebo‐treated bednets. Target concentrations of pyrethroids on bednets were permethrin 500mg/m2and lambda‐cyhalothrin 25 mg/m2. The experimental design involved random allocation of a treatment to one net per family. Whereas 68% of people questioned said they washed their nets fortnightly, observations during the 16‐week trial period showed that only 4/130 (3%) of nets involved in the trial had been washed as frequently as once per month.Early morning searches for mosquitoes under bednets (1 day/week for 16 weeks) found significantly more mosquitoes (60%An. gambiae) in placebo‐treated nets than in pyrethroid‐treated nets. The numbers found with each of the three pyrethroid treatments did not differ significantly from each other. Insecticidal efficacy of the treatments was tested by bioassays using wild‐caught unfed mosquitoes exposed to netting for 3min. Linear regression analysis of bioassay mortality against number of times that a net had been washed by villagers showed that nets impregnated with the wash‐resistant permethrin retained their insecticidal properties better than nets impregnated with lambda‐cyhalothrin or with the standa