In spite of the increasing extremity and adverse consequences of urban floods under climate change, flood characteristics in most developing countries are still largely unknown due to lack of continuous monitoring. This portentous uncertainty stances high flood risk, especially to a majority low-income urban population inhabiting the flood-prone informal settlements. Physically-based hydrologic models, whose parameters are derivable from catchment features, are often used for hydrological analysis in the ungauged regions. In this study, the Hydro-logic Engineering Centre's-Hydrologic Modelling System (HEC-HMS) was used for rainfall-runoff simulation in the upper Ngong River Basin of Kenya, draining into the Kibera informal settlement, for purposes of estimating flood peak discharges (FPDs) and direct runoff volumes (DRVs), often required for flood risk management. The Soil Conservation Service-Curve Number (SCS-CN), SCS-Unit Hydrograph (SCS-UH) and Muskingum models were used for hydrological simulation from four (4) observed and fifty-four (54) hypothetical extreme rainfalls. The results obtained indicated that a 50 mm rainfall event with a duration of 2.5 h has the potential of producing FPD and DRV of about 90m3s-1 and 1.17 x 106m3 respectively at the Ngong River confluence at Sokomoko in Kibera. On the other hand, the non-intermittent 100-year storm event of about 172 mm in depth, occurring over a period of about 4 h, potentially generated FPD of about 460m3s-1 and DRV of about 4.89 x 106m3, portending calamitous impacts in the basin. These results can offer preliminary information on potential flood character-istics, that can be used as a baseline to support local-level flood risk mitigation measures in the ungauged Ngong River Basin.