From 2003 to 2005, studies were undertaken in low altitude regions of Costa Rica and Guatemala on the economic importance for farmers' revenues of the sales of timber and fuel wood derived from trees in coffee agroforestry systems. In Costa Rica, they showed that sales of timber can account for 15 to 34% of the value of coffee revenues accumulated over a rotation period of 25 years while timber and fuel wood represented up to 52% and 25%, respectively, of the annual coffee revenues of medium size farms of low altitude in Guatemala. The commodity chains were found to be poorly organized with an absence of agreements and little cooperation among stakeholders so that farmers would benefit more by eliminating intermediaries and selling directly to wholesalers. In 2005, a study was undertaken in 51 farms from 3 cooperatives of high altitude zones (> 1200 m) of Costa Rica in order to analyse the financial profitability for farmers willing to participate in a pilot project of the "CAFE-Practices Programme" (CPP) of Starbucks. The financial analysis indicated that farmer's income was positive with a relation Benefit/Costs superior to 1, irrespective of the adoption of the CPP, for all farm types (from small farms of ~3 ha-1 relying exclusively on household labour to large farms of ~24 ha-1 with permanent labour). As the environmental requirements of the CPP were not demanding and evaluation criteria not precisely defined, their implementation did not represent a large extra cost for farmers. Therefore, results indicated that adoption of these CPP requirements improved only slightly the financial return of all farm types as coffee prices were already high in these high altitude regions renowned for their coffee quality. Consequently, this Starbucks initiative guaranteed good incomes to farmers in the medium term, encouraged them to take higher social responsibility towards their workers, but did not motivate them to fundamentally change their agricultural practices toward