This study examined the effect of poor drainage systems on the performance of urban roads in Kampala, Uganda. A quantitative, cross-sectional design was applied to an illustrative dataset of 180 simulated road segments, using the Pavement Condition Index as the main measure of performance. Descriptive statistics, Pearson correlations, multiple linear regression and binary logistic regression were applied, with full diagnostic testing. The linear model explained 82.9 percent of the variance in pavement condition. Better drainage condition and the presence of lined drains raised the index, while flooding, ponding of water on the carriageway, pavement age, traffic and heavy vehicles lowered it. In the logistic model, ponding of water multiplied the odds that a segment had required major rehabilitation in the last five years by 2.21, and each one-point improvement in drainage condition reduced those odds by about 44 percent. The study concluded that drainage was a principal determinant of pavement life in Kampala and recommended integrated planning, routine drain maintenance and the incorporation of drainage into every road project. All figures were simulated for illustration and do not describe real roads.