Background: Social media use is near-universal among adolescents and young adults, with up to 95% of those aged 13–17 reporting active use and more than one third using it "almost constantly." Anxiety and depression indicators have risen sharply in this age band since the early 2010s, raising concern about a causal contribution. Aim: This narrative review synthesizes correlational, longitudinal, and experimental evidence linking social media overuse to anxiety and depression, with attention to mediators, vulnerable populations, and intervention efficacy. Material and methods: PubMed, MEDLINE, Google Scholar, the Journal of Education, Health and Sport (JEHS), and Quality in Sport (QS) were searched for peer-reviewed publications from 2014 to 2026, supplemented by World Health Organization and U.S. Surgeon General reports. Inclusion was restricted to original research, meta-analyses, systematic reviews, and randomized trials. Results: Pooled estimates indicate that depression risk rises by 13% per additional hour of daily use, and adolescents using platforms more than three hours daily approximately double their risk of poor mental health outcomes. Effects are larger in girls than boys and are mediated principally by sleep displacement, upward social comparison, fear of missing out, and cyberbullying. Randomized trials show that limiting use to roughly 30 minutes daily, or abstaining for one week, produces measurable reductions in depressive and anxious symptoms. Critical reanalyses caution that aggregate effect sizes are small in absolute terms, with platform use explaining roughly 0.4% of adolescent well-being variance. Conclusions: Social media overuse is consistently associated with higher anxiety and depression risk, with the strongest signal in adolescent girls and in patterns marked by passive nighttime browsing. Time-limiting and abstinence interventions are promising and warrant integration into preventive mental-health care.