The effectiveness of village-wide use of permethrin-impregnated bed nets or eave, window, and door curtains as control measures for Plasmodium falciparum malaria was evaluated during two successive high-transmission seasons in western Kenya. Pairs of villages were assigned to one of three study groups: bed net, curtain, or control. Clinical, parasitologic, and entomologic measures were made from March to July 1990 and again 12 months later. When compared with the controls in 1990 and 1991, we observed a marked reduction in the incidence of P. falciparum infections in children less than six years old in the bed net villages (reduced by 40% and 48%) and a smaller but still significant reduction in the curtain villages (10% and 33%). Significant reductions were also seen in the incidence of P. falciparum parasitemias greater than 2,500/mm3 in the bed net group (reduced by 44% and 49%) and curtain group (16% and 32%). Additionally, we observed significant reductions in the incidence of documented fevers in association with P. falciparum parasitemia in bed net (reduced by 63%) and curtain villages (53%) when compared with controls. Entomologic inoculation rates in both bed net and control villages decreased by more than 50% below control values during both high transmission seasons. The results of this study, together with a 1988 study in the same area during the low transmission season, show that bed nets offer greater year-round of protection against P. falciparum infection than curtains. However, during the high transmission season, this technique reduces the frequency of P. falciparum infection rather than preventing it entirely.