The objective of this study was to determine whether house characteristics could be used to further refine the residual insecticide-spraying program in Sri Lanka. Indoor-resting mosquito densities were estimated in 473 houses based on fortnightly collections over a two-and-a-half-year period. The type of house construction and the exact location of all houses were determined. In a multivariate analysis, distance of less than 750 meters between a house and the main vector-breeding site was strongly associated with the presence of Anopheles culicifacies in the house (odds ratio [OR] 4.8, 95% confidence interval [CI] 3.4-6.8) and to a lesser extent with the presence of An. subpictus (OR 1.4, 95% CI 1.1-1.7). Poor housing construction also was an independent risk factor (OR for An. culicifacies 1.3, 95% CI 1.0-1.9; OR for An. subpictus 1.3, 95% CI 1.0-1.6). It is recommended that a malaria control strategy focus on residential areas within 750 meters of streams and rivers, with special attention given to areas with the poorest type of house construction.