This article investigates the determinants of social capital by analyzing an original survey of post-Soviet Central Asia. It tests hypotheses derived from two related questions: whether networks, norms, and trust are empirically related and the extent to which four factors—culture, regime type, perceptions of government responsiveness, and development interventions—predict levels of social capital. The results show that trust and norms diverge from networking. Interaction is higher under less repressive regimes and is further increased by development interventions; trust and norms are higher under conditions of greater repression. Culture does not affect any indicators of social capital, but perceptions of responsiveness correlate with higher levels of trust. As such, disaggregating social capital is a promising new direction for research. Scholars should investigate why the components of social capital sometimes correlate but at other times diverge, and they should consider the possibility of distinct causal mechanisms in their development.