variavel0=Cibele Vieira Falkenberg - cvfalkenberg@hotmail.com UIUC Nísio de Carvalho Lobo Brum - nisio@serv.com.ufrj.br COPPE/UFRJ Abstract. Abstract. The objective of the present work is to simulate thermal comfort for humans exposed to dynamic and non-uniform boundary conditions according to a scale similar to that established by ASRHAE (1997). Fiala’s (1998) model equations and properties for passive and active systems were used and the model was integrated numerically using the finite difference method. According to the Fiala model the human body is compartmentalized in ten different parts, each consisting of up to seven tissues (brain, bone, lung, core, muscle, fat, skin). The parts were assigned specified radiation and convection coefficients, and were thermally coupled by perfusion. The results are organized in terms of three case studies. The first case study concerns thermal comfort under two metabolic rates in rooms with a positive displacement air conditioning system. For the same room temperature, two different temperature and humidity gradients, and uniform air temperature were compared. The second case simulated thermal comfort of passengers in Rio de Janeiro subway trains. The passenger was initially exposed to 35o C air temperature and 55% relative humidity, followed by exposure to two different air temperature and humidity conditions for the next 30 minutes. The third case corresponds to thermal comfort of workers under high metabolic rates (3.5 met) in a hangar of the Macapá thermoelectric unit. Indoor air temperature was around 40o C, 30% relative humidity. Workers were initially acclimated to outdoor air conditions (~30o C, 78%). The simulation considered three air velocities and 30 minutes exposure. In all cases, the extended multi-compartment model proved to give reasonable predictions for human thermal comfort sensations, as well as providing guidance for improved air conditioning design. Keywords. Thermal Comfort, indoor air distribution, dynamic thermal conditions.