This study examined several tenets of the response styles theory of depression (RST Nolen-Hoeksema, 1991) and its generalizability to adolescent populations. Using a short-term longitudinal design, we evaluated whether response styles (i.e., ruminative and distractive responses to depressed mood) predict depression severity and whether they do so beyond the effects attributable to other cognitive variables. Anxiety was also assessed to investigate the specificity of RST to depression. Finally, we investigated gender differences in response style and whether these differences account for the gender difference in depression that emerges during adolescence. Self-reported affective symptoms and cognitive predictors (attributional style and private self-consciousness) were assessed in 397 adolescents. Results indicated that response style variables predicted concurrent and future (6-week followup) mood scores beyond effects accounted for by other cognitive variables. Rumination was less specific to the prediction of depression than was distraction, but only rumination predicted future depressed mood. Although girls reported more rumination, controlling for response style did not eliminate the gender difference in depression. Results are discussed with respect to the current status of RST, including construct measurement and its relevance to adolescent depression.