The optimal management of water resources requires that the collected hydrogeological, meteorological, and spatial data be simulated and analyzed with appropriate models. In this study, a catchment-scale distributed hydrological modeling approach is applied to simulate water stress for the years 2000 and 2050 in a data scarce Pra Basin, Ghana. The model is divided into three parts: The first computes surface and groundwater availability as well as shallow and deep groundwater residence times by using POLFLOW model; the second extends the POLFLOW model with water demand (Domestic, Industrial and Agricultural) model; and the third part involves modeling water stress indices—from the ratio of water demand to water availability—for every part of the basin. On water availability, the model estimated long-term annual Pra river discharge at the outflow point of the basin, Deboase, to be 198 m 3 /s as against long-term average measurement of 197 m 3 /s. Moreover, the relationship between simulated discharge and measured discharge at 9 substations in the basin scored Nash–Sutcliffe model efficiency coefficient of 0.98, which indicates that the model estimation is in agreement with the long-term measured discharge. The estimated total water demand significantly increases from 959,049,096 m 3 /year in 2000 to 3,749,559,019 m 3 /year in 2050 (p \ 0.05). The number of districts experiencing water stress significantly increases (p = 0.00044) from 8 in 2000 to 21 out of 35 by the year 2050. This study will among other things help the stakeholders in water resources management to identify and manage water stress areas in the basin. Keywords PCRaster Á GIS Á River discharge Á Water vulnerability Á Groundwater Á POLFLOW