AbstractThermal comfort is very important in any work or operation environment. But “thermal comfort” is a very vague and not easily defined term, and it is influenced by both the physical environment and the individual’s physiology or psychology. To at least partially overcome these problems, this work proposes the use of a fuzzy adaptive network (FAN) to model the thermal comfort system. To illustrate the approach, actual experimental data were used to train the network and to give results. Although only very simple examples were used, the results show the usefulness of the proposed approach.