Background: Tuberculosis (TB) incidence in peadtrics and children living with human immune-deficiency virus (HIV) is an emerging global concern. Although, the incidence of TB among adult HIV patients is exhaustively studied in Ethiopia, but among children on HIV/AIDS care is overlooked. Knowledge of the time when TB develops during successive follow up could be helpful for time relevant intervention strategies.Methods: health institution based retrospective cohort study conducted among 421 children on HIV/AIDS from 2009-2018. Time to develop TB was defined as time from enrollment for ART care until development of TB among children on ART. Proportional hazard assumption was checked for each variable and no variable was found with Schoenfeld test <0.05. Variables with P-value <0.25 at bivariate Cox regression analysis were entered into multivariable Cox model. Multivariable Cox regression model with 95%CI and AHR was used to identify significant predictor variables to develop TB at P< 0.05.Result: Totally 421 children were followed for a total of 662.5 Person Years of observation (PYO). The maximum and minimum follow up time on ART was 0.37 and 4.49 years, respectively. The median age of the children on ART at enrollment was 8 years (IQR=2-15). The Overall incidence density of tuberculosis in HIV infected children was 9.6/ 100 PYOs 95%CI (8.06-10.3). Tuberculosis occurrence among HIV infected children was significantly associated within TB history of contact AHR=3.7, 95%CI (2.89-7.2), not started on cotrimoxazole(CPT) AHR=2.4: 95%CI (1.84-4.74), incomplete vaccination AHR=2.4, 95%CI (1.32-4.5), sever stunting AHR =2.99:95%CI (1.2-7.81), having hemoglobin (Hgb) ≤10 mg/dl AHR = 4.02, 95%CI (2.01-8.1).Conclusion: More than 80% of TB incidences occurred during two years of follow up after ART started. So intensified screening of CPT& therapeutic feeding is highly recommended for all children.