Storage tanks that are underutilized and have poor mixing characteristics represent a significant potential for water quality degradation within the distribution system. Computational fluid dynamic (CFD) modeling and water temperature measurements can be very effective tools to predict mixing characteristics inside the tanks. This paper focuses on the use of CFD modeling and temperature measurements to determine mixing characteristics in storage tanks, using a case study that occurred with the City of Virginia Beach. The results from mixing predictions are verified by water quality data and used to identify the storage tanks with inadequate mixing characteristics and thus the greatest potential for water quality degradation. Once the nature of the mixing problem is identified, it becomes easier to make the appropriate physical and operational modifications to improve mixing in the storage tanks. This paper also presents general recommendations for improving mixing, and reports on actual physical modifications that have improved mixing characteristics in a storage tank. Includes figures.