Post-dispersal seed predation plays an important role in plant demography and biodiversity maintenance. However, the effects of seed density on seed predation from previous studies have been inconsistent. We dissected the effects of density on the 2-step processes of seed predation using 101 520 seeds from 62 plant species in an alpine pine forest for 3 consecutive years. In this study we explained the current controversy surrounding the effects of density on seed predation. Seed encounter frequency (at least 1 seed being predated from an experiment depot) showed positive density dependence, while seed exploitation (the proportion of seeds being predated of the encountered depots) showed negative density dependence. Both density effects showed a consistent trend but with different magnitudes of effect across years. Final seed predation is the combination of seed encounter and seed exploitation. Final seed predation could be either positively or negatively density-dependent and was contingent on the magnitude of the difference between positive density-dependent seed encounter and negative density-dependent seed exploitation. Our results also indicated that studies including only a few species would produce biased results, because the density effect on seed predation differed greatly among plant species. Future studies should include a large number of plant species that possess a wide range of diverse seed traits to avoid potential bias and produce more comprehensive and accurate results.