In a recent meeting of IEA's Annex 23, several members presented their conclusions on the modeling of phase change materials behavior in the context of building applications. These conclusions were in agreement with those of a vast review, involving the survey of more than 250 journal papers, undertaken earlier by the group of Ecole de technologie superieure. In brief, it can be stated that, at this point, the confidence in reviewed models is too low to use them to predict the future behavior of a building with confidence. Moreover, it was found that overall thermal behaviors of PCM are poorly known, which by itself creates an intrinsic unknown in any model. Models themselves are most of time suspicious as they are often not tested in a very stringent or exhaustive way. In addition, it also appears that modeling parameters are somewhat too simplified to realistically describe the complete physics needed to predict the real life performance of PCMs added to a building. As a result, steps are now taken to create standard model benchmarks that will improve the confidence of the users. Hopefully, following these efforts, confidence will increase and usage of PCM in buildings should be eased.