--- - |2+ Abstract Objective To examine the interaction effect of anxiety and depression on the intentional forgetting of positive and negative valence words. Methods One hundred fifty‐five grade 7 to grade 10 students participated in the study. The item‐method directed forgetting paradigm was used to examine the intentional forgetting of positive‐valence, negative‐valence, and neutral‐valence words. Results Negative‐valence words were recognized better than either positive‐valence or neutral‐valence words. The results revealed an anxiety main effect (p = .01, LLCI = −.09, and ULCI = −.01) and a depression main effect (p = .04, LLCI = .00, and ULCI = .24). The anxiety score was negative, whereas the depression score was positively related to the directed forgetting of negative‐valence words. Regression‐based moderation analysis revealed a significant anxiety × depression interaction effect on the directed forgetting of positive‐valence words (p = .02, LLCI = .00, and ULCI = .01). Greater anxiety was associated with more directed forgetting of positive‐valance words only among participants with high depression scores. With negative‐valence words, the anxiety × depression interaction effect was not significant (p = .15, LLCI = − .00, and ULCI = .01). Conclusion Therapeutic strategies to increase positive memory bias may reduce anxiety symptoms only among those with high depression scores. Interventions to reduce negative memory bias may reduce anxiety symptoms irrespective of levels of depression. - Journal of Clinical Psychology, Volume 74, Issue 9, Page 1509-1525, September 2018.