Recent evidence calls into question the perfect information assumption underlying use of the Deferred Acceptance Algorithm (DAA) in school matching. We study performance of the DAA when learning is costly for students, by introducing a matching model in which schools agree on their rankings of students, each student’s prior is exchangeable across schools, and learning costs are linear in Shannon mutual information. We characterize the unique equilibrium outcome analytically for any number of schools and students, any exchangeable priors, and heterogeneous marginal costs of learning. We demonstrate how each student’s rank, learning costs and prior beliefs interact to determine gross and net welfare and the specific mistakes they make. We show how policies that lower costs of learning may reduce inequities.