When decision-making in the case of imminent water-related humanitarian disasters or violent conflict depends on a quick assessment of water-related risks, there may not be enough time to collect high-quality local data. Online available global data may offer an alternative data source. We present a method to construct a water resources model based on global datasets. We apply it to the Upper Niger Basin in West Africa and test its credibility with hydrological performance metrics and ‘fit for purpose’ indicators. The ‘fit for purpose’ indicators are tailored to questions the model should support answering and allow assessment of whether a better-performing model would result in a different basis for decision-making. We find that the model scores ‘satisfactory’ on both types of indicators. Thus, models based on global open datasets may be suitable for rapid, preliminary assessments where time is of the essence or high-quality local data is unavailable.