Abstract This meta-analysis examines the effect of cognitive-behavioral therapy (CBT) for depressive symptoms among college students. Studies were identified through searches in five electronic databases (PubMed, EMBASE, Web of Science, PsycINFO and Cochrane Central). We included studies examining the effect of CBT compared with control conditions, alternative interventions, and multi-modal interventions that combined cognitive-behavioral and other therapeutic approaches. We combined effect sizes (Hedge's g) using random-effects models at post-intervention and three follow-up assessments. 20 trials (n = 2185) met our inclusion criteria, resulting in 23 comparisons between a CBT condition and a comparison condition. At post-intervention, CBT was found more effective than control conditions (g = 0.48; 95% CI: 0.31, 0.66). In particular, CBT administered to high-risk students produced a larger effect size than those administered to low-risk students. The effect of CBT was largely maintained at 6 months follow-up (g = 0.52; 95% CI: 0.06, 0.99), but not at 12 months follow-up (g = 0.09; 95% CI: −0.10, 0.27). There were few indications that CBT was more or less effective than alternative interventions. We did not find adding other components to CBT produced stronger effects than CBT alone. Despite few indications for the superiority of CBT to other interventions, CBT could be still proposed as a promising candidate for addressing depressive symptoms among college students in the short term, particularly for those endorsing higher risk. Further research applying rigorous methods is needed to confirm and extend the findings of this meta-analysis.