We use the household production framework to theoretically connect sociability and purposive incentives for volunteering and two forms of social capital: social connections and civic capacity. Then, using a unique statewide data set, we estimate the determinants of (a) the probability of receiving social capital benefits and (b) the level of such benefits. We show that: religious and social service organizations have a large impact on social capital formation; the probability of being socially and civically engaged increases with volunteering; and two-adult families are more likely to feel socially and civically engaged. These results are consistent with recent aggregate evidence on the decline of social capital in the United States: social capital formation declines with less religious and altruistic orientation at the community level, and as families move away from a two-adult family structure. By contrast, through volunteering, one can increase the likelihood of being socially and civically engaged.