We examined adhesive seed dispersal, recruitment and site occupancy of 17 plant species characteristic of semi-natural grasslands. The main objectives were to examine different aspects of dispersal: the potential distance of seed transport; the likelihood of recruitment at putative "suitable sites" along road verges; the spatial (realised) patterns of occurrence along road verges (occupancy); and associations among these aspects and seed size. Adhesive dispersal capacity was investigated experimentally using cattle. Recruitment was investigated by seed sowing experiments. Occupancy was recorded in two regional surveys. Results suggested that adhesive dispersal may be an effective mode of seed transport, even for species which lack special "adhesive" structures, such as hooks or hairs. Fifteen of the 17 species successfully recruited at the experimental gaps and adjacent grassy areas. Disturbance promoted seedling recruitment. Seedling emergence was negatively associated with adhesive dispersal capacity. Spatial (occupancy) patterns were not related to recruitment features or dispersal capacity. There is some support for a trade-off between adhesive dispersal capacity and recruitment ability, but no support for a trade-off between adhesive dispersal capacity and ability to withstand competition during recruitment. The results are discussed in the context of species distribution and persistence in fragmented landscapes. We suggest that adhesive dispersal has been underestimated as a mechanism of seed transport, but that adhesive dispersal capacity per se is not the main factor determining species occupancies in fragmented grasslands.