Sharpshooters (Hemiptera: Cicadellidae: Cicadellinae) are potential vectors of Xylella fastidiosa, the bacterial causal agent of coffee "crespera" disease. This study examined the effects of local (shade composition, organic or conventional management, and weed management) and landscape (surrounding land uses) context on sharpshooter communities in coffee agroforestry systems (CAFS) within the Volcánica Central-Talamanca Biological Corridor (VCTBC), in Costa Rica. On a broader perspective, this study also sought to evaluate how the ecosystem service of natural pest control overlaps with other ecosystem services provided by CAFS and how decision makers view these services. The abundance and species richness of sharpshooters were evaluated in four CAFS types present in the VCTBC, using yellow sticky traps and Malaise migration traps in 2005 and 2006. Approximately 73 species of sharpshooters were found in CAFS within the VCTBC. Sharpshooters were more abundant from June to September and declined afterwards. Sharpshooter communities were distinct in each agroforestry system, and coffee sharpshooter communities were more similar to pasture than to forest sharpshooter communities. Shade was a key local variable, negatively affecting sharpshooter abundance. Weed management did not affect sharpshooter communities. Sharpshooters were more abundant at edges of coffee farms, and appeared to move from forests to coffee farms. Surrounding area in forest was positively correlated to sharpshooter abundance. Surrounding area in fallow was positively correlated to species richness. Overlapping of GIS layers of ecosystem services provided by CAFS revealed that there is more land of the VCTBC where natural pest control overlaps with watershed services than with biodiversity services. Interview data showed that natural pest control was more important for decision makers than all other services, except connectivity. Decision makers value ecosystem services in the planning of corridor activities, in addition to location and organization level of corridor areas. Although research is needed on the epidemiology of X. fastidiosa, we found that species that have tested positive for the presence of the bacteria are negatively affected by shade. Increasing shade could be a strategy for management of these species in coffee agroforestry systems. This practice also increases the level of other ecosystem services provided by coffee agroforestry systems.