Abstract Recruitment limitation (seed limitation and establishment limitation) has been proposed to contribute to the maintenance of high diversity of plants in tropical forests. This hypothesis suggests that diversity is maintained by slowing down competitive exclusion. In this study we: (1) quantified the relative importance of seed limitation and establishment limitation on recruitment for 4–7 species, testing the impact of seed predators and competition; (2) evaluated the effect of recruitment limitation on the diversity of seedling communities: if recruitment limitation is a diversity‐promoting mechanism, then removing the limitation should have a negative effect on diversity. To evaluate the magnitude and effects of recruitment limitation we used data from 400 seed traps, we executed seed addition experiments with and without vertebrate exclusion, and we conducted seedling transplant experiments during 2 yr in an Amazonian rain forest in Colombia. On the basis of seed traps, we found that seed limitation was strong; however, seed addition experiments showed that establishment limitation is more important than seed limitation for recruitment in all species considered. The addition/transplant experiments did not affect diversity of the surrounding seedlings, lending no support to the theory that recruitment limitation affects diversity. Establishment limitation was stronger at the seed–seedling transition than in older seedling transitions, as shown by high survival in the seedling transplants experiment. We found that seed dispersal increases regenerating plant diversity, and that predation is an important factor affecting seed establishment. We highlight the role of animals in these processes and therefore in the maintenance of diversity in tropical rain forests.