As membership in traditional civic organizations declines in the United States (Putnam, 2000), could volunteering for nonprofit organizations be an alternative source of social capital formation? We use an updated household production framework (Becker, 1996) to theoretically connect volunteering with two forms of social capital: social connections and civic capacity. Using a unique statewide data set from Vermont, we then use the Cragg (1971) model to estimate the determinants of the probability of receiving a social capital benefit, and the level of such a benefit. We first show that the probability of receiving a social connection or a civic capacity benefit from one's most important nonprofit organization is increased: (a) if it is a religious or social service organization; (b) if one increases their volunteering for the organizations; and (c) if one is female, college educated or in a two-parent family. However, the relative magnitude of volunteering is similar, or relatively small, compared to the other significant determinants. We then show that an increase of volunteer hours does increase the levels of social connection and civic capacity, but the magnitude of this effect is also relatively small.