Aims: Problematic social media use is increasing in China and could be a risk factor for depression. We investigated cross-sectional associations between problematic social media use and depressive outcomes among Chinese college students with potential mediation by perceived social support, social media violence, and loneliness. Thereafter, we evaluated the effectiveness of a one-month group counseling intervention in reducing depressive symptoms related to social media addiction. Methods: Depressive symptoms, social media addiction, perceived social support, social media violence, and loneliness were self-reported among 21,000 college students in Shaanxi province, China. A randomized controlled trial was designed based on the results of the observational study and Satir Transformational Systemic Therapy (STST) among 60 college students assigned to intervention (N = 30) or control/no treatment (N = 30). Self-administered surveys were completed at baseline (T1), at the end of the 1-month intervention (T2), and at 2-month follow-up post-intervention (T3). Results: After controlling for relevant covariates, more problematic social media use was associated with more depressive symptoms (β = 0.18, p < 0.001) and depression (OR = 1.08, 95% CI: 1.075, 1.092). Multiple mediation analyses found that perceived social support, social media violence, and loneliness significantly mediated associations between problematic social media use and depressive symptoms (model fit: RMSEA = 0.065, GFI = 0.984, CFI = 0.982). Bootstrapping revealed significant indirect effects of problematic social media use on depressive symptoms through the mediators named above (0.143, 95% CI: 0.133, 0.156). The subsequently informed intervention significantly reduced depressive symptoms at T2 (mean difference: −12.70, 95% CI: −16.64, −8.76, p < 0.001) and at T3 (mean difference: −8.70, 95% CI: −12.60, −4.80, p < 0.001), as well as levels of social media addiction, perceived social support, social media violence, and loneliness. Conclusions: Problematic social media use is a risk factor for depressive outcomes among Chinese college students, and perceived social support, social media violence, and loneliness mediate this association. STST-based group counseling may reduce depressive symptoms related to high social media usage in this population.