Objectives:Retaining on antiretroviral therapy is essential for reducing HIV-related morbidity and mortality. However, attrition in HIV-positive children remains a critical challenge in resource-limited settings, including Ethiopia. This study aims to determine the incidence and predictors of attrition among children on antiretroviral therapy at the University of Gondar Comprehensive Specialized Hospital, Northwest Ethiopia.Methods:An institution-based retrospective follow-up study was conducted among 357 HIV-positive children at the University of Gondar Comprehensive Specialized Hospital from 1 January 2005 to 30 December 2018 (G.C.). Data were collected by chart review using a structured and pre-tested data abstraction checklist. SPSS 22 and STATA 14.0 were used for data entry and analysis, respectively. In the Cox proportional hazard model, bivariables had a 0.25 computed to multivariable, and variables with a p-value of 0.05 at a 95% confidence interval were considered statistically significant predictors of attrition incidence.Results:A total of 344 child records with a completeness rate of 96.4% were reviewed and included in the analysis. The median follow-up period was 4.3 (interquartile range = 4.3 ± 4.7) years, and the median survival time was 132 months. The incidence rate of attrition was 6.6 per 100 person year observation (PYO) (95% confidence interval = 5.5, 8.0). In all, 105 (30.5%) children were recorded as attrition in the follow-up period. Baseline WHO clinical staging 3 and 4 (adjusted hazard ratio = 2.3 (95% confidence interval = 1.3, 4.0)), baseline weight-for-age −2 Z-score (adjusted hazard ratio = 3.1 (95% confidence interval = 1.7, 5.3)), cotrimoxazole non-users (adjusted hazard ratio = 2.5 (95% confidence interval = 1.4, 4.3)), and baseline hemoglobin levels 10 mg/dL (adjusted hazard ratio = 2.7 (95% confidence interval = 1.5, 47)) were found to be a predictor of attrition.Conclusion:The overall incidence of the rate of attrition was high. Baseline WHO clinical staging 3/4, baseline hemoglobin 10 mg/dL, cotrimoxazole (cotrimoxazole preventive therapy) non-user, and underweight at baseline (weight-for-age 2 Z-score) were found to be the main predictors of attrition.