Colonisation theory predicts that large-seeded species are more likely to show increased seedling recruitment in response to seed addition than are small-seeded species. This is both because their seedlings tend to have better survivorship potential, and because their background density of germinating seedlings tends to be lower. We tested this hypothesis by combining data from a recent review of seed addition studies with seed mass data. Logistic regressions showed positive relationships between seed mass and propensity to increase seedling establishment in response to seed addition in experiments in which establishment success was assessed within 6 months or a year, but not in experiments in which establishment success was assessed more than a year after seed addition. When data for all time periods were combined, a generalised linear model including terms for seed mass, time and an interaction term showed a significant positive relationship between seed mass and species response to seed addition. Thus, knowing a species' seed mass significantly increased our ability to predict its response to seed supplementation.