Background: Peripheral intravenous cannula is venous access devices widely practiced 30% to 80% of admitted patient requiring intravenous therapy. The failure rate is high in infants. This will cause prolonged hospital stays, it seeks additional medical care costs, painful procedure, and frequent cannulation attempts adversely affecting the person’s overall hospital experience. Even though the problem is existing, there is limited information in Amhara region as well as in Ethiopia. So, this study provides valuable information. Objective: To assess time to failure of peripheral intravenous cannula and predictors among infants admitted in public hospitals of Bahir Dar City, Northwest Ethiopia, 2022. Methods and materials: An institution based prospective follow up study was conducted among 378 infants admitted to neonatal intensive care unit and pediatric ward at public hospital of Bahir Dar City Northwest Ethiopia from May 8, 2022 to June 22,2022. Systematic random sampling technique was employed. The data was entered Epi data 4.6 version and exported to STATA version 14 software. Log rank test was used to test for the presence of difference among predictor variables. Data quality was assured. Outcome variables were dichotomized into censored or failure median failure time and incidence rate was calculated. Variables with p-value <0.25 were included in multivariable and p-values < 0.05 in the multi-variable Cox regression analysis were considered as significant predictors of peripheral intravenous cannula failure. Result: From a total 382 sampled infants with peripheral intravenous cannula, 378(98.9%) were included in the final analysis. Of those participants, 281 (74.3%) had a cannula failure, yielding an incidence rate of 15.2 per 1000 person-hours with median time 46 hours. In the final model, peripheral intravenous canula failure, gestational age <32 weeks (AHR: 2.09, 95%, CI=1.44,3.03), flushing/irrigation by saline at insertion time (AHR: 0.58, 95%, CI=0.44,0.75), insertion site at leg (AHR: 1.5, 95%, CI=1.14,2.03), not visible venous (AHR: 1.41, 95%, CI=1.11,1.79) and 4-6 times attempts insertion (AHR: 1.26, 95%, CI=1.52,3.37) of were statically significant predictors of peripheral intravenous cannulas failure. Conclusion: The overall incidence of peripheral intravenous failure was high. Gestational age < 32 weeks, cannulation on leg, not visible vein, 4-6 times insertion attempts, and flushing/irrigation by saline were significant predictors for peripheral intravenous cannula failure. Key words: peripheral intravenous, cannula failure, infants, neonates