The relationship between adult attachment (anxious attachment and avoidant attachment) and problematic social media use is controversial and contradictory. Hence, the present study clarified the relationship between adult attachment dimensions and problematic social media use through meta-analysis. To better explain the relationship, this study also established meta-analytic structural equation modeling to examine the mediating role of social anxiety in the relationship. This study used CMA and R software for data analysis. Forty-five effect sizes were included in the study, including 11,746 individuals. Results showed that anxious attachment was strongly correlated with problematic social media use (r = 0.319, 95 %CI[0.271, 0.366]), whereas avoidant attachment was weakly correlated with problematic social media use (r = 0.091, 95 %CI[0.011,0.170]). Moderating effects showed that the relationship between anxious attachment and problematic social media use was moderated by the measurement instrument. Meta-analytic structural equation modeling showed that anxious attachment had a significant positive effect on PSMU through social anxiety, anxious attachment had a significant positive effect on PSMU through social anxiety. This study clarifies inconsistencies in the existing literature through meta-analysis, providing reliable conclusions and novel perspectives. It assists clinical practitioners in developing tailored treatment programs for practical interventions. It is suggested that treatment for problematic social media use requires attention to individuals with insecure adult attachment and social anxiety. For individuals with high anxious attachment and social anxiety, it is essential to help them manage their social media use effectively and reduce their dependence on it. Concurrently, interventions for attachment avoidant individuals should focus on enhancing their social self-confidence to reduce the influence of social anxiety on their social media use.