The potential of low impact development (LID) practices in stormwater management has received increasing attention. However, the impacts of antecedent dry-weather periods (ADP) and rainfall magnitude on LID performance in the mitigation of urban flooding and non-point pollution remain largely unexplored. In this study, the LID performance under different ADP and rainfall patterns was investigated using a hybrid hydraulic and water quality modeling approach with consideration of uncertainty and sensitivity. The results showed that LID practices mitigated the runoff volume, peak flow, and non-point pollutant loads by a maximum of 57.16%, 52.46%, and 76.27%, respectively. An increase in the length of the ADP improved the LID performance by 43.98%–77.25%, while an increase in rainfall magnitude weakened the LID performance by 29.55%–73.43%. The uncertainty in the model caused a relatively greater impact on LID water quality performance than on LID hydraulic performance, and the uncertainty slightly increased with an increase in ADP length and rainfall magnitude. The ADP length had a positive impact on LID performance, and the rainfall magnitude had a negative impact on LID performance. In addition, the LID performance was heterogeneously distributed in space, and the spatial heterogeneity was exacerbated by both a decrease in ADP length and an increase in rainfall magnitude. The methods and findings presented in this study provide an additional reference for sustainable stormwater management by investigating the LID performance and influencing factors.