The amount and distribution of foliar feeding injury by adult potato flea beetles, Epitrix cucumeris (Harris), were examined on individually caged potato plants grown in field plots in Manitoba. Plants were either maintained as uninfested controls, or were exposed throughout the growing season to different insect densities that mimicked the natural seasonal pattern of infestation. In 1984, a trial was conducted using cv. ‘Norland’ exposed to four different densities of potato flea beetles. In 1989 and 1990, cv. ‘Russet Burbank’ was exposed to potato flea beetles, and in some treatments, plants were exposed to early summer infestations of Colorado potato beetles, Leptinotarsa decemlineata (Say). In each trial, during the late summer period of high potato flea beetle density, the amount and distribution of flea beetle feeding injury were assessed at weekly intervals. Counts of feeding punctures in single leaflets were made from leaves in the upper, middle, and lower third of each caged plant, and these data were subjected to repeated measures analysis of variance. In each of the 3 years, increasing the number of flea beetles increased the mean number of feeding punctures per leaflet in an approximately linear fashion; however, the number of punctures per beetle varied between cultivars and years. In 1984 and 1990, the number of feeding punctures per leaflet was least in the upper third of the plants, and greater in the lower, or middle and lower, third of plants. However, in 1989, the vertical distribution of feeding punctures was relatively even. Previous feeding by Colorado potato beetles increased the mean number of flea beetle feeding punctures per leaflet and changed the vertical distribution of feeding punctures. Rainfall and temperature were correlated with patterns of flea beetle injury; injury was concentrated on lower leaflets during weeks of greater rainfall, and upper leaflets were injured most during weeks with higher average temperatures. It is concluded that flea beetles exhibit preferences for feeding in specific portions of potato plants, and that these preferences change in response to previous defoliation and are influenced by meteorological conditions. Consequently, counting feeding punctures would not be a reliable method of assessing whether control measures for potato flea beetles are justified.