Heat insulation applied on outer wall surfaces of buildings for the purpose of conserving energy, can be analyzed experimentally, mathematically and by using simulation modelling. In this study, simulation modelling of insulation layer (d(2)), for residential buildings in 81 cities classified into four climatic regions in Turkey was investigated. This stimulus model is constructed into rule-based Mamdani-type fuzzy modelling (RBMTF), using input parameters (U(w),T(e)) and output parameter d(2), described by RBMTF if-then rules. RBMTF has been designed using the MATLAB 7.04 fuzzy logic toolbox. The d(2) was predicted using two input parameters: U(w) and T(e). The values for U(w) ranged from 0.4 to 0.7. The values for T(e) ranged from -43 degrees C to -3 degrees C. The values for d(2) ranged from 0.07 to 0.21. Seven linguistic terms were considered for each of the input and output variables, namely Very Low (L1), Low (L2), Negative Medium (13), Medium (L4), Positive Medium (L5), High (L6), Very High (L7). R(2) for the testing data were about 97.4%. Overall, RBMTF can be used as a reliable modelling method for thermal performance of multi-layer precast concrete panels used in residential buildings' studies. (C) 2010 Elsevier Ltd. All rights reserved.