Purpose: The purpose of this study was to determine if collective student trust functions as a resource for urban elementary students. Methods: Data from 1,646 students nested in 56 elementary schools in an urban school district were used to test the hypothesized effect of collective student trust on school identification, self-regulated learning, and math and reading achievement. A model-building process in HLM 6.08 was used to test the three hypotheses. Random-intercepts means-as-outcomes models were used to assess the school-level effects on school identification and self-regulated learning, controlling for free and reduced lunch rate and prior achievement. A one-way ANCOVA with random intercepts was used for math and reading achievement. Free and reduced lunch was specified as a student- and school-level control in order to set a more conservative standard for detecting a trust effect. Results: The hypothesized effects of collective student trust on school identification, self-regulated learning, and achievement were confirmed. A culture of student trust in urban elementary schools partly contributed to identification with school, internal control over learning tasks, and math and reading achievement. Collective trust was the strongest school-level antecedent of positive student beliefs, behavior, and achievement. Implications: With policy debates centering on increased funding, accountability, and teacher and leader evaluation, trust may seem like an unlikely mechanism to ameliorate performance problems in urban schools. Results from this study suggest that policy makers and school leaders should not overlook student trust as a viable social resource for schools and students.