Cognitive bias modification of interpretations (CBM-I) programs, in which individuals are trained to interpret ambiguous scenarios in a benign way, appear effective in altering anxiety-related cognitive biases in both children and adults. In this experimental study, we explored the effectiveness of a novel CBM-I training tool for children, which involves joint discussions of ambiguous information with a same-gender peer. 10- to 11-year-old boys and girls (n = 20) were provided with ambiguous social vignettes, each followed by two interpretations, and then asked to select one of them after a brief discussion with a same-gender peer. A further group of participants did not participate in any training but only completed pretraining and posttraining measures (n = 18). Results indicated that children who completed the interpretation training made less negative interpretations, endorsed less negative emotional consequences, reported less social anxiety, and performed better in a stressful task compared with the no-intervention group. Clinical implications of the results are briefly discussed.