Urban road drainage systems in conflict-affected cities represent one of the most critically underexamined components of post-war infrastructure rehabilitation. This study presents a comprehensive, data-driven failure analysis of urban road drainage systems in Juba City, South Sudan, integrating hydrological field surveys, EPA SWMM v5.1 stormwater modelling, Intensity–Duration–Frequency curve development, and structured engineering interviews across 12 major arterial roads surveyed systematically between 2020 and 2024. A total of 148 drainage failure incidents were catalogued and classified into four primary failure modes: hydraulic overloading (73.0%), operational blockage (18.2%), structural deterioration (6.1%), and systemic design defects (2.7%). Pareto analysis confirms hydraulic overloading as the singular dominant failure mechanism. Rainfall IDF curves derived from 35 years of Juba International Airport meteorological station data, fitted to the Gumbel Extreme Value Type I distribution with regression coefficients a = 1847.3, b = 0.201, c = 14.7, e = 0.812 (R² = 0.94), reveal that the existing drainage network operates at a mean hydraulic capacity of only 37% of that required under the 25-year return period standard recommended for major urban arterials. Manning’s equation analysis of 47 surveyed cross-drain sections confirms a mean capacity deficit of 2.3 m³/s (SD = 1.87 m³/s), with 87.2% of sections exhibiting measurable underperformance. EPA SWMM simulation for the critical Gudele Ring Road catchment (138.7 ha) achieved a Nash–Sutcliffe Efficiency of 0.89, validating the hydrological modelling approach and predicting surface flooding depths of 0.34–1.12 m persisting for 4.2 hours post-storm peak. A GIS-based phased rehabilitation framework integrating composit