Understanding dispersal systems is a key step towards improving our ability to predict species migration trajectories in response to climate change. However, the abiotic seed dispersal systems of tropical plants remain largely overlooked, with studies inferring seed dispersal from fruit morphology than conducting empirical tests. Here, we determined seed dispersal season experimentally tested the seed dispersal mode, and estimated dispersal distance kernels for five edaphic specialists, mountaintop endemics from campo rupestre, a tropical, megadiverse and nutrient-impoverished grassland. We monitored fruiting phenology for 12 months and established field and laboratory experiments to assess the role of abiotic factors in releasing seeds from dry capsules and to measure seed dispersal distance. Contrary to previous studies that suggested abiotic dispersal, we found no evidence of water- or wind-operated primary seed dispersal mechanisms for all species. We show that fruiting peaks coincided with optimum conditions for seedling establishment, and that short seed dispersal distances are consistent with the idea of reduced dispersibility in edaphic specialists. Seed buoyancy experiments show a large potential for secondary seed dispersal by water run-off. Our data suggest that abiotic dispersal in the study species plays a limited role in displacing seeds away from parental plants, a pattern that markedly contrasts with that found in other vegetation types. Short-distance dispersal in campo rupestre is consistent with theory explaining the evolution of phylomatry, but it also represents reason for concern of species migration in face of global climate changes. The combination of edaphic endemism and limited dispersal in mountaintops species provides limited opportunity for migration and poses a significant threat to their long-term survival.