Introduction: Internalized HIV stigma is a key risk factor for negative outcomes amongst adolescents living with HIV (ALHIV), including non-adherence to anti-retroviral treatment, loss-to-follow-up and morbidity. This study tested a theoretical model of multi-level risk pathways to internalized HIV stigma among South African ALHIV.Methods: From 2013 to 2015, a survey using total population sampling of ALHIV who had ever initiated anti-retroviral treatment (ART) in 53 public health facilities in the Eastern Cape, South Africa was conducted. Community-tracing ensured inclusion of ALHIV who were defaulting from ART or lost to follow-up. 90.1% of eligible ALHIV were interviewed (n = 1060, 55% female, mean age = 13.8, 21% living in rural locations). HIV stigma mechanisms (internalized, enacted, and anticipated), HIV-related disability, violence victimization (physical, emotional, sexual abuse, bullying victimization) were assessed using well-validated self-report measures. Structural equation modelling was used to test a theoretically informed model of risk pathways from HIV-related disability to internalized HIV stigma. The model controlled for age, gender and urban/rural address.Results: Prevalence of internalized HIV stigma was 26.5%. As hypothesized, significant associations between internalized stigma and anticipated stigma, as well as depression were obtained. Unexpectedly, HIV-related disability, victimization, and enacted stigma were not directly associated with internalized stigma. Instead significant pathways were identified via anticipated HIV stigma and depression. The model fitted the data well (RMSEA = .023; CFI = .94; TLI = .95; WRMR = 1.070).Conclusions: These findings highlight the complicated nature of internalized HIV stigma. Whilst it is seemingly a psychological process, indirect pathways suggest multi-level mechanisms leading to internalized HIV stigma. Findings suggest that protection from violence within homes, communities and schools may interrupt risk pathways from HIV-related health problems to psychological distress and internalized HIV stigma. This highlights the potential for interventions that do not explicitly target adolescents living with HIV but are sensitive to their needs.