Learning objectives (LOs) are statements that typically precede a study session and describe the knowledge students should obtain by the end of the session. Despite their widespread use, limited research has investigated the effect of LOs on learning. In three laboratory experiments, we examined the extent to which LOs improve retention of information. Participants in each experiment read five passages on a neuroscience topic and took a final test that measured how well they retained the information. Presenting LOs before each corresponding passage increased performance on the final test compared with not presenting LOs (experiment 1). Actively presenting LOs increased their pedagogical value: Performance on the final test was highest when participants answered multiple-choice pretest questions compared with when they read traditional LO statements or statements that included target facts (experiment 2). Interestingly, when feedback was provided on pretest responses, performance on the final test "decreased," regardless of whether the pretest format was multiple choice or short answer (experiment 3). Together, these findings suggest that, compared with the passive presentation of LO statements, pretesting (especially without feedback) is a more active method that optimizes learning.