The overall output of migrants and the numbers migrating different distances from a habitat patch was investigated by pitfall trapping Collembola in a patchy habitat. Tussocks of Carer ursinae constituted the preferred habitat for Collembola. The tussocks were embedded in bare soil with a cover of cyanobacteria, and distributed with varying distances between them. In a previous study, it was found that Collembola show species-specific patterns of abundance and demographic composition relative to the degree of isolation of tussocks. It is hypothesized that species differences in migration rate, mobility range and colonization success are partly responsible for these distribution differences. In two series of experiments, pitfall traps were used to investigate the migration of Collembola out of tussocks. The captures were analyzed in relation to species, population size in experimental tussock and distance from tussock. This study shows that there are clear differences between species in the output of migrants from populated tussocks, and these differences probably contribute to the observed species distribution patterns in relation to habitat patchiness. Only Hypogastrura viatica showed a consistent positive relationship between population density and output of migrants. Hypogastrura viatica and H. longispina also had the highest overall output of migrants, and their high mobility can thus counteract the effects of patch isolation. Migration ability only seems to limit the distribution of one of the species (Onychiurus groenlandicus). Only a few individuals of O. groenlandicus were captured outside populated tussocks in this study, and the species is generally not present in the most isolated tussocks. Two Folsomia species (F. sexoculata and F. quadrioculata), appeared to have very similar migration abilities, although they have completely different distribution patterns with respect to habitat patchiness. In addition to the investigations on migration, trials on passive transport were done through net trapping on surface water from snowmelt. This appeared to be a mode of dispersal which primarily affects surface living species, but did not carry large numbers of migrants and is probably not of great importance.