Most of the cocoa production worldwide comes from small farms where cocoa is family-grown, often on highly diversified and small plots known as agroforests. In the Dominican Republic, most cocoa is produced in such systems where undescribed cocoa cultivars are associated with more than 40 other plant species. These highly diversified and shaded cocoa orchards are the current strategy of a highly vulnerable and ageing population of local farmers with low income and poor education level. As most of the next generation of farmers has already fled to the cities or has found a job in the flourishing tourism sector, there is an urgent need for solutions to make cocoa production attractive again. Government's strategies for cocoa intensification traditionally rely on genetic improvement and capacity building on crop, pests and disease management, both having proved to fail because of a lack of knowledge about actual farmer's practices and strategies. On the other hand, a recent alliance between Research and chocolate manufacturers has led to an innovative program for the design of cocoa-based agroforestry systems. The design methodology relies on participatory approach based on the identification, quantification and ranking of the products generated by the current coca-based AFS. The farmer's strategy of diversification where timber and mostly fruit trees are cultivated with cocoa trees on the same plot, together with other annual and multiannual crop, is the basis of the participatory approach. Workshops with focus groups are conducted to deal with every aspect of cocoa production, from soil quality management to cocoa and shade plants management. These workshops lead to define a panel of options, or prototypes, where cocoa varieties are associated with a productive shade canopy linked to markets, in a design that allow an acceptable cocoa yield level together with other sources of income. The panel of agroforestry prototypes produced in these participatory workshops is f