Direct topical impact of an insecticide spray on a population of a nontarget leaf-eating beetle, Gastrophysa polygoni, was studied, and the relative importance of phenology, instar susceptibility and instar speciÐc exposure was evaluated. Two insecticides, cypermethrin and dimethoate, were investigated. In the laboratory, topical toxicity to eggs, second-instar larvae and adults was recorded in dose-response experiments. The spatial distribution of larvae and eggs were measured in the Ðeld. Deposition of insecticide onto eggs, secondinstar larvae and adult specimens was measured at di †erent positions within the crop canopy by use of a dye tracer technique. A temperature-driven population model was constructed to simulate population development of all life stages in the Ðeld. The model was based on laboratory measures of growth and development at various temperatures. Mortality due to direct insecticide exposure was calculated as a function of population demography, spatial distribution of individuals, spatial deposition of the insecticide, and stage-speciÐc susceptibility. Cypermethrin had the greatest impact, reducing population size by 19È32%. The life stages most sensitive to cypermethrin were the larval instars. As the population developed from eggs to larvae and imagines, the impact of one spraying Ðrst increased and then decreased according to the proportion of larvae in the population. Dimethoate had less e †ect on the population, i.e. 1É9È7É6% reduction. Dimethoate was most toxic to the egg stage, and consequently the e †ect on the population decreased as the proportion of eggs decreased due to hatching. The direct e †ect of insecticide spraying was signiÐcantly a †ected by all three factors investigated, i.e. phenology, life stage susceptibility and stage-speciÐc exposure. The latter factor is composed of both spray Ñux at various spatial positions in the canopy and the ability of di †erent life stages to retain spray droplets.