Five previously developed temperature-distribution models are compared using data from actual thermal energy storage (TES) systems, and the advantages and disadvantages of each model are identified. These models are used for evaluating the energy and exergy in a TES. Furthermore, the accuracy and simplicity of the temperature-distribution models (Linear, Stepped, Continuous-linear, Three-zone and General-linear) are compared in order to determine which model has the greatest accuracy and simplicity regarding computational effort, while providing physical understanding. After fitting the models to temperature data for 35 stratified TES samples, it was observed that some models are easy to use but relatively inaccurate, while others are relatively accurate but complicated to use. As a trade-off, the Three-zone model provides physical understanding, yields acceptable accuracy and is relatively simple to use, especially in computing the parameters needed for energy and exergy assessments of stratified TESs.