Direct-seeding trials in Tasmania tested the effectiveness of five treatments: a clear polymer film removed at three time intervals; insecticide; wetting agent; irrigation; and caging to exclude all mammal herbivores. The most effective treatment in improving early establishment of 15 native species across 2 years and at five sites was the application of the insecticide Bifenthrin to reduce seed predation by ants. The polymer film removed after 10 weeks also improved germination of a range of shrubby but not tree species. However, these early advantages were eroded over time at the first trial site due to damage from herbivores and at subsequent sites due to weed competition and mortality of seedlings over summer. Caging was beneficial to the ongoing persistence of tree and shrub species but also protected weeds. Further work on seed testing and timely weed control is needed.