This review examined evidence for some core predictions of the response styles theory (RST) concerning the relation between response styles and symptoms of depression and gender differences in the use of response styles in non-clinical children and adolescents. In summarizing the literature, effect sizes (pooled correlation coefficients) were calculated for cross-sectional and longitudinal studies. Stability of the obtained effect sizes was evaluated by means of a fail-safe N analysis. Results indicated that stable and significant effect sizes were found for rumination being associated with concurrent and future levels of depression. When controlling for baseline levels of depression, effect sizes for rumination and distraction were not stable, indicating that these findings should be interpreted with considerable caution. Finally, significant and stable effect sizes for gender differences in response styles were found only for rumination among adolescents. Taken together, the findings partly support the predictions of the response styles theory examined in this meta-analysis and may implicate that rumination is a cognitive vulnerability factor for depressive symptoms among adolescents.