PURPOSE To test whether an academic detailing intervention would improve journal reading among residents. METHOD In 1995-96, the authors conducted a randomized controlled trial at two family medicine residencies in Missouri. Fifty-nine family practice residents were randomized into two groups identical in baseline characteristics; 29 received individual 15-minute educational presentations, based on academic detailing, that emphasized careful selection of medical journal articles for reading. The authors measured the percentage of important journal articles of which the residents were aware, the percentage of those they had read (either abstract only or the whole article), and the percentage of correctly answered knowledge questions. RESULTS Despite randomization, the residents in the control group indicated on the pretest that they were more likely than were the residents in the intervention group to be aware of and to have read abstracts from selected articles. However, when comparing pretest with posttest results, the intervention group recalled 18.2% (95% CI, 2.0, 34.0) more articles and correctly answered 16.6% (95% CI 3.0, 29.9) more knowledge questions than did the control group. CONCLUSION A brief intervention increased residents' recall of important articles and knowledge of those articles' conclusions.