Populations at range margins may be genetically different from more central ones for a number of mutually non-exclusive reasons. Specific selection pressures may operate in environments that are more marginal for the species. Genetic drift may also have a strong effect in these populations if they are small, isolated and/or have experienced significant bottlenecks during the colonisation phase. The question if peripheral populations are special, and if yes then how and why, is of obvious relevance for speciation theory, as well as for conservation biology. To evaluate the uniqueness of populations at range margins and the influence of gene flow and selection, we performed a morphometric study of two grassland butterfly species: from Swedish populations that are peripheral and isolated from the main area of the species distributions and from populations in the Baltic states that are peripheral but connected to the main area of the species distributions. These samples were compared to those from central parts of the species distributions. The isolated populations in both species differed consistently from both peripheral and central populations in their wing size and shape. We interpret this as a result of selection caused by differences in population structure in these isolated locations, presumably favoring different dispersal propensity of these butterflies. Alternative explanations based on colonisation history, latitudinal effects, inbreeding or phenotypic plasticity appear less plausible. As a contrast, the much weaker and seemingly random amongregion differences in wing patterns are more likely to be ascribed to weaker selection pressures allowing genetic drift to be influential. In conclusion, both morphological data and results from neutral genetic markers in earlier studies of the same system provide congruent evidence of both adaptation and genetic drift in the isolated Swedish populations of both species.