Changes in coffee economics are leading producers to reduce agrochemical use and increase the use of shade. Research is needed on how to balance the competition from shade trees with the provision of ecological services to the coffee. In 2000, long-term coffee experiments were established in Costa Rica and Nicaragua to compare coffee agroecosystem performance under full sun, legume and non-legume shade types, and intensive and moderate conventional and organic inputs. Coffee yield from intensive organic production was not significantly different from intensive conventional in Nicaragua, but in Costa Rica it was lower during three of the six harvests. Full sun coffee production over 6 years was greater than shaded coffee in Costa Rica (61.8 vs. 44.7 t ha -1 , P = 0.0002). In Nicaragua, full sun coffee production over 5 years (32.1 t ha -1 ) was equal to coffee with shade that included Tabebuia rosea (Bertol.) DC., (27-30 t ha -1 ) and both were more productive (P = 0.03) than coffee shaded with Inga laurina (Sw.) Willd. (21.6 t ha -1 ). Moderate input organic production was significantly lower than other managements under all shade types, except in the presence of Erythrina poepiggina (Walp.) O.F. Cook. Inga and Erythrina had greater basal area and nutrient recycling from prunings than other shade species. Intensive organic production increased soil pH and P, and had higher K compared to moderate conventional. Although legume shade trees potentially provide ecological services to associated coffee, this depends on management of the competition from those same trees.