Water resources planning and management is crucial and challenging in semi-arid regions to minimize water scarcity. Potential impacts due to climate change are a concern to water managers and stakeholders in semi-arid river basins with limited water availability. This study provides a probabilistic assessment of climate change impacts on water scarcity in the Sevier River Basin of Utah, which has a snowmelt-driven water supply and high agricultural water demands, using a decision-scaling framework. The methodology consists of a bottom-up approach that uses climate response functions, together with projections from 31 general circulation models (GCMs), to assess vulnerability to water scarcity for 2000–2099. Water scarcity is defined using an index comparing water availability to crop water demand predicted by the AquaCrop model from the Food and Agriculture Organization. Results showed that off-season precipitation is the most sensitive factor affecting water scarcity in the basin, followed by precipitation and temperature during the growing seasons. The GCM projections of temperature and precipitation suggest an increasing availability of water for agriculture in the basin. Still, a considerable risk probability of agricultural water shortage was found in years 2025 through 2049 with the emission scenario RCP4.5, suggesting the need for adaptation and mitigation strategies. The bottom-up decision scaling approach used here with a wide range of GCMs was practical to explore climate risk to agricultural water scarcity given the simplicity and minimal computational requirement.