In an examination of the effects of timing of feedback on student learning, participants were randomly assigned to complete five introductory course tests using either control (Scantron form) or one of three feedback (end-of-test, 24-hr delay, immediate) procedures. A cumulative final examination with 50 new items and 10 items repeated from each course test was used to assess retention during the academic semester. Fifty final examination items, administered at postcourse intervals of 3, 6, 9, and 12 months, were used to measure longer-term retention. The provision of feedback at either the end of a test or after a 24-hr delay, relative to control procedures, increased retention for 3 to 5 weeks during the semester. Robust enhancement of retention at all intervals was observed only when immediate feedback was provided. The magnitude of the immediate feedback effect indicates that students benefit from affirmation of correct responding and notification of incorrect responding during classroom tests and suggests noteworthy potential for creating learnable moments during which students validate and self-correct their knowledge stores. (Contains 6 figures.)