Hawai‘i has served as a model system for studies of nutrient cycling and conservation biology. The islands may also become a laboratory for exploring new approaches to forest restoration because of a common history of degradation and the growing number of restoration projects undertaken. Approximately half of the native ecosystems of Hawai‘i have been converted to non-native conditions. Many restoration projects have focused on intensively managed out plantings of native plants with emphasis on threatened and endangered species. While these projects have been effective in stabilizing plant populations, this model is often prohibitively expensive for restoration at the scale needed to protect watersheds and provide habitat for rare bird species. Here we suggest ways of rethinking ecological restoration that are applicable across the tropics, particularly on islands and fire-prone grasslands. First, we suggest making use of non-native, non-invasive species to help reclaim degraded or invaded sites or as long-term components of planned restoration outcomes. Second, we suggest incorporating remote sensing techniques to refine where restoration is carried out. Finally, we suggest borrowing technologies in plant production, weed control, and site preparation from industrial forestry to lower restoration costs. These suggestions would result in ecosystems that differ from native reference systems in some cases but which could be applied to much larger areas than most current restoration efforts while providing important ecosystem services. We also stress that community involvement is key to successful restoration, as a major goal of almost all restoration projects is to re-connect the community with the forest.