This paper looks at the potential for poverty alleviation in one part of Uganda, based on a poverty analysis of the local, and on analyses of the local civil society and of development discourses that are often dominated by the central over the local. In response to calls for microstudies of actually existing civil society it points to the usefulness of including community wide processes and hegemonic discourses in analyses of the local civil society’s development role. Inspired by the IDS civil society Concept paper’s list of civic organizations’ roles: “ • representation of the interest of specificic groups in relation government and to other sectors of society; • mobilization of social actors to increase their conciousness and impact; • regulation and monitoring of state performance and the behaviour and actions of public officials; • developmental or social action to improve the wellbeing of their own or other constituencies,” (IDS 1998, p6) It includes both the great plurality of local groups found in Rubaya, a poor, fairly isolated and distant sub-county in Southwest Uganda, as well as community planning and plan implementation, as important parts of civil society analyses. Rubaya is really poor compared to other parts of Uganda, especially because of its constrained non-agricultural sector, and because people seem to be making up for that by selling food crops to get an income, which to some extent endangers their own food security. It is also clear, however, that there is social differentiation with repercussions for different groups of farmers’ agricultural and livelihood strategies, as well as their abilities to respond to agricultural interventions. The analysis revealed that Rubaya is endowed with a rich and vibrant local civil society, which is a potential link or counterforce to larger external institutions – developmental or financial in nature. A civil society that does mirror the social differentiation, but which still, despite their somewhat lower group membership and planning participation, encompasses the poorer sections of society. In the last 15-20 years 2-4 different community planning processes have been introduced into and adopted by village communities, which have proven to be able to plan and implement work on their own needs as communities. Participatory methods can make it more inclusive, as in the CARE FIP project, but disintegration into disparate planning processes, may lead to community frustration. New civil society structures, in ther form of local functional groups have also mushroomed in recent years, because of new opportunities, partly opened by NGO interventions like CARE FIP. They enhance further the basis for agricultural innovation – which also, contrary to expectations, appear to have been existing for a long time. Contrary to the popular development discourse about Kabale district, farmers have historically invented and/or adopted basic additions to, and thus maintained their farming-systems, in response to population growth, new socio-economic conditions, and externally introduced innovations. While this is supported by mutual communication and being organized through local groups, and by community capacities being enhanced through participatory planning processes, both NAADS and FIP, as new extension initiatives, are, though inclined in that direction, to different degrees in their conception and implementation far from really embracing a new paradigm , where most important would be empowering farmers on technology development, by enabling them to: identify problems; work with problem solutions; experiment with their own and new solutions; – and thus enhance their self respect, also by showing that outsiders respect, what they are doing.