The paper responds to a recent article by Shashua-Bar et al. that asserted that the modification proposed by Elnahas and Williamson to the original formulation of the "cluster thermal time constant" (CTTC) model for predicting air temperature in the urban canopy layer causes serious errors in predictions. It reviews the development of both versions of the model, highlighting the differences between them. A methodology is suggested for analysis of the quality of the model predictions and the performance of the Elnahas-Williamson version is evaluated. The analysis shows that both versions of the CTTC model give similar predictions in stable meso-climatic conditions, but the Elnahas-Williamson version is clearly superior in changing weather. The revised CTTC model is proposed as the basis for a tool to account for urban modification to air temperature in the simulation and design of HVAC systems in buildings. (c) 2005 Elsevier B.V. All rights reserved.