Community building is a key ingredient in comprehensive community initiatives(CCI) to improve poor neighborhoods throughout the United States but the scholarly literature evaluating the results of this approach is sparse. CCI refers to the activities of local organizations, governments, religious institutions, or other agents to increase leadership, resident participation, and effective cooperation to achieve shared goals. CCIs emerged in the late 1980’s in the United States as responses to several trends: 1) the decline of the welfare state and the devolution of responsibility for the welfare of inner city residents and communities to local government and the non-profit sector 2) recognition of the interdependence of problems of physical deterioration, economic disinvestment, social marginality, crime rates, and levels of political influence for distressed inner city neighborhoods;3)leadership of liberal foundations who believed that by being comprehensiveness, coordinated, and collaborative with local institutions, and involving community participants, successful interventions could be achieved. All draw on the idea of social capital. Relationships that lead to effective cooperation constitute the social capital of the community, which can be used to achieve collective and individual goals. This paper uses three theories of social capital to evaluate a community building initiative in four small neighborhoods in Harlem and the Bronx, and focus on housing. These different frameworks critically draw on three competing ideas about social capital: 1) theories that focus on the role of voluntary associations and intervening institutions in democratic societies (Putnam,2000); 2) a functionalist theory of social capital, emphasizing the utility of social networks for the achievement of individual and group goals (Coleman, 1988); and 3) Bourdieu’s (1986) concept of social capital as the social process reproducing the class structure of capitalism and controlling individual access to capital and credit through class membership. Bourdieu reminds us of three significant issues for would- be community builders: 1) the absence of social capital that translates into good educations, jobs, and a wealth of other positive outcomes is actively produced through exclusionary practices that are integral to the structure of mainstream institutions 2) efforts to generate social capital readily replicate the “fault lines” of society such as race, class, age, and gender ; 3) if social capital is to be useful for achieving capital intensive goals such as better housing, then new relationships must be built between the wealthier sectors of society and low-income, minority communities (Briggs, 1997). Most CCI’s try to address these issues by developing relationships with local institutions to improve access and inclusion, by attempting to hire staff who reflect the composition of the community, by self-conscious outreach to different sectors of the community, and by engaging government and the private sector in community building efforts. Nonetheless, problems of access to networks and institutions rich in economic and social capital limit the achievements of CCIs. DeFilippis (2001) argues that community development efforts attempting to build social capital divorced from economic capital ultimately contribute little to the lives of the poor and powerless. He particularly faults Putnam’s use of social capital as a property of a community as misleading because it denies the unequal power relations and access to economic capital that exist in a community and between particular communities and those who possess power and capital in the larger society. Housing provides a tough test for the use of social capital to improve disadvantaged communities because good housing requires the investment of financial capital as well. This paper evaluates whether social capital contributes to better housing directly and how, if at all, it leverages the investment of financial resources in that housing. Looking across the four sites, several variables emerge as particularly relevant: 1) population resources; 2) the initial condition of the housing stock; 3) the success CBOs have in building social capital; 4) different ownership forms; and 5) the strategies CBOs use to bring pressure on owners of deteriorated housing for improvement. Single site analysis have indicated that the part of housing quality related to behavior is very responsive to increases in social capital through CCIs, but the part related to financial investments are more complex. Housing markets in these communities also varied in ways that affected the success of housing improvement strategies. This paper will us quantitative and qualitative data, HLLM and GSI, for the four sites to examine these issues.