As soon as it has been introduced, at the turn of the century in Sumatra and West Kalimantan (Indonesia), rubber has been a major deforestation agent (2.5 millions hectares in 1998), mostly smallholdings (83% of the total rubber area). However, these rubber low input complex agroforestry systems, "jungle rubber", may be considered as agents of reforestation as there are currently the main reservoir of biodiversity in regions where forest has almost disappeared, The "jungle rubber" helped reversing the idea of a tree-crop being a pure deforestation agent. Since the 1980s, governemental projects have introduced technical change and "modern agriculture" based on clonal material through monocropping systems. Although the main target was to increase the Indonesian supply of rubber, the theory claiming that technological progress in agriculture reduced pressure on forests may have helped to promote these monoculture systems. However, some contraints limit the development of the monoculture system : shortage of labour, lack of information, nonavailability of improved planting material as well as lack of credit, in particular when projects stop their activities due to governemental disengagement. If the conditions of a pioneer phase are gathered (abundance of land, reservoir of labour, a crop opportunity and an attractive sustainable market), it is true that technological progress is likely to increase deforestation and traditionnal land-use. Technological progress may reduce the pressure on forest only when the pioneering effect is already slowing down. The rubber showcase helps to demonstrate it. One major technological progress is the recent use of herbicides to control Imperata cylindrica on degraded lands. Many rubber farmers reintroduce also agroforestry practices in former monoculture plots or develop their own systems when not under influence of extension. In that case, technological progress clearly seems to encourage reforestation through the development of agrofo