The hypothesis that corridors increase animal movement between habitat fragments, a central tenet of conservation biology, has been virtually untested. This study demonstrates that corridors increase interpatch movement rates of two butterfly species. The hypothesis was tested in a large-scale, replicated experiment, in which 27 equal-sized (1.64-ha) patches of early successional habitat were created within large areas of pine forest. Patches varied in whether or not they were connected to another patch by a corridor, and in their distance from other patches (64–384 m). The results of mark–release–recapture studies showed that two open-habitat butterfly species, Junonia coenia and Euptoieta claudia, moved more frequently between patches connected by corridors than between unconnected patches. Interpatch movement was significantly, negatively related to interpatch distance. Interpatch movement rates of J. coenia were significantly, positively related to the density of its host and nectar resource, Linaria canadensis. Corridor effects were stronger for males than for females and were most pronounced within three days after butterflies were marked. Pine forest was not a complete barrier to butterfly movement; both species moved between unconnected patches, even at the longest distances. However, the results of this study suggest that corridors will increase long-distance movements of habitat-restricted species.