Used meta-analysis to examine the efficacy of bibliotherapy. Bibliotherapy treatments were compared to control groups and therapist-administered treatments. The mean estimated effect size (d) of the 70 samples analyzed was + 0.565. There was no significant differences between the effects of bibliotherapy and therapist-administered treatments, as well as no significant erosion of effect sizes at follow-up. Bibliotherapy did appear more effective for certain problem types (assertion training, anxiety, and sexual dysfunction) than for others (weight loss, impulse control, and studying problems). Overall the amount of therapist contact during bibliotherapy did not seem to relate to effectiveness, but there was evidence that certain problem types (weight loss and anxiety reduction) responded better with increased therapist contact. Recommendations for future research were given, especially for more research on the commonly purchased books and moderator analyses by personality type and reading ability.