Considerable reforestation was undertaken around the world in the second half of the twentieth century. Nonetheless, very large areas of degraded land and forest still remain present across the tropics. Indeed, these areas continue to increase. Several countries do have experience in undertaking large-scale reforestation including Japan, Korea, China and Vietnam. There are lessons to be learned from their experiences but not simple recipes for others to follow. Ecological and socio-economic differences mean that those undertaking reforestation must adjust their approaches to suit the circumstances present in particular locations. The task is made even more difficult because the conditions under which future reforestation might be undertaken are changing. For example, there is uncertainty about the availability of land (because of the need for increased food production), the impact of urbanization, the future markets for forest products and ecosystem services and about how to establish forests able to adapt to a changing climate. Given this background, several policy issues need resolution if reforestation is to be undertaken on a large scale. These include deciding (i) how much reforestation should be undertaken in particular landscapes, (ii) just where in these landscapes this should be done, (iii) the types of reforestation to be carried out in these different areas, (iv) who should make these decisions, and (v) how reforestation should be implemented. All of this means there is a rich field of study for silviculturalists wishing to look beyond establishing simple monocultures of fast-growing exotic species. Silvicultural systems designed for industrial timber plantations are not necessarily those suited to overcoming forest and land degradation where a variety of stakeholders are involved and where these stakeholders are interested in the provision of ecosystem services as well as (or instead of) goods such as timber. It also means that foresters will have to bridge the divide between the natural or physical sciences and the social sciences rather better than has been done in the past. I will highlight some of the key questions (though not necessarily the answers) that I think deserve more attention than they have received hitherto.