BACKGROUND: Patients who are lost to follow-up while on treatment compromise their own health and the long-term success of antiretroviral therapy (ART) programs. Besides, loss to follow-up (LTFU) increases HIV-related morbidity and mortality. Therefore, this study aimed to determine the incidence of LTFU and its predictors among adult HIV positive patients on anti-retroviral therapy at North Shoa zone public hospitals, Northeast Ethiopia. METHODS: A retrospective follow up study of 517 people living with HIV/AIDS and attending an ART clinic between 2015 and 2020 was conducted at North Shewa zone, public hospitals. Kaplan-Meier failure function together with log rank test was used to compare failure function. Multivariable Cox proportion hazards regression model was used to determine predictors of LTFU. RESULT: The incidence density rate of lost to follow up among HIV positive adult on ART was found to be 8.9 per 100 adult years observation (95%CI; 7.45, 10.68). In multivariable cox proportional regression analysis, WHO clinical stage-IV (AHR = 1.50; 95% CI: 1.08, 3.75), comorbidity disease (AHR = 0.54; 95% CI; 0.30, 0.97), body mass index less than 18kg/m2 (AHR = 1.60; 95% CI; 1.02, 2.51), cotrimoxazole preventive therapy (AHR = 1.57; 95% CI;1.09, 2.53), and a low CD4 count (AHR = 1.66; 95% CI; 1.29, 3.49) were found to be a significant predictors of lost to follow up. CONCLUSION: The current study showed that the incidence rate of loss to ART follow-up was high. Body mass index score less than 18kg/m2, advanced WHO clinical stage, CD4<200cell/mm, had comorbidity disease, and cotrimoxazole therapy were a significant predictors of lost to ART follow up. Therefore, appropriate mitigation measures in the at-risk group need to be instigated to advance retention rate.