The majority of insect species do not show an innate behavioural migration, but rather populations expand into favourable new habitats or contract away from unfavourable ones by random changes of spatial scale. Over the past 50 years, the scienti¢c fascination with dramatic long-distance and directed mass migratory events has overshadowed the more universal mode of population movement, involving much smaller stochastic displacement during the lifetime of the insects concerned. This may be limiting our understanding of insect population dynamics.In the following synthesis, we provide an overview of how herbivorous insect movement is governed by both abiotic and biotic factors, making these animals essentially`slaves of their environment'. No displaced insect or insect population can leave a resource patch, migrate and £ourish, leaving descendants, unless suitable habitat and/or resources are reached during movement. This must have constrained insects over geological time, bringing about species-speci¢c adaptation in behaviour and movements in relation to their environment at a micro-and macrogeographical scale. With insects that undergo longrange spatial displacements, e.g. aphids and locusts, there is presumably a selection against movement unless overruled by factors, such as density-dependent triggering, which cause certain genotypes within the population to migrate. However, for most insect species, spatial changes of scale and range expansion are much slower and may occur over a much longer time-scale, and are not innate (nor directed).Ecologists may say that all animals and plants are ¢guratively speaking`slaves of their environments', in the sense that their distribution is de¢ned by their ecology and genotype. But in the case of insects, a vast number must perish daily, either out at sea or over other hostile habitats, having failed to ¢nd suitable resources and/or a habitat on which to feed and reproduce. Since many are blown by the vagaries of the wind, their chances of success are serendipitous in the extreme, especially over large distances. Hence, the strategies adopted by mass migratory species (innate pre-programmed £ight behaviour, large population sizes and/or fast reproduction), which improve the chances that some of these individuals will succeed. We also emphasize the dearth of knowledge in the various interactions of insect movement and their environment, and describe how molecular markers (protein and DNA) may be used to examine the details of spatial scale over which movement occurs in relation to insect ecology and genotype.