Objective: This study aimed to investigate the associations between perceived social support, self-esteem, depressive symptoms, and internet addiction among university students, providing theoretical and empirical foundations for the prevention and intervention of internet addiction in this population. Methods: In September 2023, a cross-sectional survey was conducted using convenience sampling to recruit 1141 university students. Validated instruments, including the Perceived Social Support Scale (PSSS), Internet Addiction Test (IAT), Self-Esteem Scale (SES), and Beck Depression Inventory (BDI), were administered to collect data. Statistical analyses, including correlation analysis and mediation effect testing, were performed using SPSS 26.0 and AMOS 26.0. Results: Perceived social support significantly negatively predicted internet addiction (β = −0.437, P < 0.001), with a direct effect accounting for 39.55% of the total variance. Mediation analysis revealed that self-esteem and depressive symptoms independently mediated the relationship between perceived social support and internet addiction, with indirect effects of 43.98% and 13.21%, respectively. Furthermore, a significant serial mediation pathway through self-esteem to depressive symptoms was identified, contributing 3.26% of the total effect. Conclusion: The findings highlight robust associations between perceived social support, self-esteem, depressive symptoms, and internet addiction severity among university students. Interventions targeting the enhancement of social support networks, fostering self-esteem, and mitigating depressive symptoms may effectively reduce the risk of internet addiction in this population.