Pesticides, in particular insecticides, can be very beneficial but have also been found to have harmful side effects on non-target insects. Butterflies play an important role in ecosystems, are well monitored and are recognised as good indicators of environmental health. The amount of information already known about butterfly ecology and the increased availability of genomes make them a very valuable model for the study of non-target effects of pesticide usage. The effects of pesticides are not simply linear, but complex through their interactions with a large variety of biotic and abiotic factors. Furthermore, these effects manifest themselves at a variety of levels, from the molecular to metapopulation level. Research should therefore aim to dissect these complex effects at a number of levels, but as we discuss in this review, this is seldom if ever done in butterflies. We suggest that in order dissect the complex effects of pesticides on butterflies we need to integrate detailed molecular studies, including characterising sequence variability of relevant target genes, with more classical evolutionary ecology; from direct toxicity tests on individual larvae in the laboratory to field studies that consider the potentiation of pesticides by ecologically relevant environmental biotic and abiotic stressors. Such integration would better inform population-level responses across broad geographical scales and provide more in-depth information about the non-target impacts of pesticides.