The control and the monitoring of indoor thermal conditions represents a pre-eminent task with the aim of ensuring suitable working and living spaces to people. Especially in industrialised countries, in fact, several rules and standards have been recently released in order of providing technicians with suitable design tools and effective indexes and parameters for the checking of the indoor microclimate. Among them, predicted mean vote (PMV) index is often adopted for assessing the thermal comfort conditions of thermal moderate environments. Unfortunately, the PMV index is characterised by non-linear features, that could determine some difficulties when monitoring and controlling HVAC equipment. In order of overcoming these problems, a fuzzy control for HVAC system is here described. It represents a new simple approach, focused on the application of an adaptive fuzzy controller that avoids the modelling of indoor and outdoor environments. After a brief description of the method, some simulation results are presented. A simplified application, referring to a room belonging to an university building, is finally reported. (C) 2003 Elsevier B.V. All rights reserved.