Background and Objectives: Growing literature documents that where you live has an impact on your health, due in part to social capital. Building on social capital literature, we assess how subjective appraisals of neighborhood quality are associated with self-reported health (SRH) for older adults. Research Design and Methods: Cross-sectional analysis of the 2014 California Health Interview Survey, a representative survey of diverse, noninstitutionalized California residents. We use three measures of neighborhood quality: trustworthy neighbors, helpful neighbors, and feeling safe. Using weighted ordinary least squares regression, we assess the associations of trust, helpfulness, and safety to SRH, controlling for neighborhood, demographic, and health care variables. We then examine how these associations vary by household income. Results: We find that characterizing neighbors as helpful and feeling safe are associated with better SRH, even controlling for community, demographic, and health care variables. However, the importance of these dimensions varies across household income: helpfulness is positively associated, whereas trust is negatively associated with SRH for lower income residents; safety is positively associated with SRH in all but the lowest income residents. These findings show that social capital dimensions work differently from one another, and differentially affect the health of older adults. Discussion and Implications: Scholarly analyses of neighborhood effects should include a range of social capital measures and stratify by household income. Our findings may also inform priority setting for social capital programs, especially for older adults with limited economic resources. Policies and programs should consider actions that raise perceptions of helpfulness and safety.