This paper reports on the application of a monthly rainfall-runoff model for the Okavango River Basin. Streamflow is mainly generated in Angola where the Cuito and Cubango rivers arise. They then join and cross the Namibia/Angola border, flowing into the Okavango wetland in Botswana. The model is a modified version of the Pitman model, including more explicit ground and surface water interactions. Significant limitations in access to climatological data, and lack of sufficiently long records of observed flow for the eastern sub-basins represent great challenges to model calibration. The majority of the runoff is generated in the wetter headwater tributaries, while the lower sub-basins are dominated by channel loss processes with very little incremental flow contributions, even during wet years. The western tributaries show significantly higher seasonal variation in flow, compared to the baseflow dominated eastern tributaries: observations that are consistent with their geological differences. The basin was sub-divided into 24 sub-basins, of which 18 have gauging stations at their outlet. Satisfactory simulations were achieved with sub-basin parameter value differences that correspond to the spatial variability in basin physiographic characteristics. The limited length of historical rainfall and river discharge data over Angola precluded the use of a split sample calibration/validation test. However, satellite generated rainfall data, revised to reflect the same frequency characteristics as the historical rainfall data, were used to validate the model against the available downstream flow data during the 1990s. The overall conclusion is that the model, in spite of the limited data access, adequately represents the hydrological response of the basin and that it can be used to assess the impact of future development scenarios. Keywords: Okavango basin, Rainfall-runoff models, Regional calibration.