Many climate datasets are currently available for predicting buildings' thermal-energy performance in dynamic simulation environment. The most used weather files are the Typical Meteorological Year (TMY) and Test Reference Year (TRY). The purpose of this research is to experimentally highlight the need for a frequent update of such weather files used when predicting buildings performance in urban areas in order to include microclimate phenomena such as UHI and temperature extremities. To this aim, a statistical comparison between the data collected by two urban weather stations in central Italy and the data from the TMY and TRY weather files of the same area is carried out. The results highlight a summer UHI of 5.5 degrees C and -3.5 degrees C in the late afternoon and early morning, respectively. Additionally, the comparison of TMY and TRY data with those ones collected by the urban stations shows non-negligible seasonal discrepancies in terms of the main microclimate monitored parameters, i.e. dry bulb temperature, relative humidity, solar radiation, and wind. Such inconsistencies question the reliability of traditional weather files to be used in building dynamic simulations to accurately estimate needs and production within the city energy grid. (C) 2017 Elsevier Ltd. All rights reserved.