Purpose: The purpose of this study is to test the validity of using collective trust as a social indicator of instructional capacity. Design/methodology/approach: A hypothesized model was advanced for the empirical investigation. Collective trust was specified as a latent construct with observable indicators being principal trust in faculty (PTF), faculty trust in principal (FTP), faculty trust in colleagues (FTC), and faculty trust in students (FTS). It was hypothesized that enabling school structure is directly related to the latent collective trust construct and collective trust is directly related to school performance. Data were collected in the spring of 2010/11 from teachers and students in 85 schools in an urban school district in a southwestern state. A partially latent structural regression model was tested in AMOS 7.0. Findings: Results of the measurement model support the theoretical relationship among faculty trust in principal, faculty trust in colleagues, faculty trust in students, and principal trust in faculty. Both directional hypotheses were supported: enabling school structure had a strong, direct effect on a culture collective trust and collective trust had a strong, direct effect on school performance. Research limitations/implications: The sample consisted of schools in one urban district in the southwestern part of the USA, and collective trust only operationalized the social dimension of instructional capacity. Practical implications: Regular and consistent measures of collective trust have the potential to improve how administrators at site and district levels manage the implementation of improvement strategies designed to build capacity. Originality/value: Many theoretical discussions treat trust as a constitutive property of capacity building, but few studies have empirically tested a "priori" models that specify relationships among structures and processes aligned with instructional capacity, collective trust, and school performance. (Contains 5 tables and 5 figures.)