This paper describes newly developed model-in-the-loop simulation (MILS) which makes design for cooperative climate control logics between automotive HVAC (heating, ventilation and air conditioner) and auxiliary thermal devices more efficient in considering thermal sensation and comfort of occupants. The auxiliary thermal devices such as an air-conditioned seat and a heated steering wheel consume less energy than the HVAC, and they have a potential to improve the total energy consumption satisfying thermal comfort of occupants. However, it is not easy to design the cooperative climate control logics for these thermal devices since thermal sensation and comfort must be taken into account while the logic optimization. The proposed MILS consists of thermal-environment model, thermal-device model, ambient conditions and climate control logics. The thermal-environment model simulates dynamics of temperature distribution in a cabin and human body. Compartment model is adopted for the thermal environment model, and it provides higher-speed simulations than distributed parameter model. Thermal sensation and comfort of occupants are computed on the basis of heat loss of occupants. Local thermal sensation of occupants are estimated by equivalent temperature (Teq), and we confirmed that Teq had a linear correlation to declared values of subjects. UC Berkeley's comfort model is used to convert local sensation to overall sensation and comfort. An HVAC, an air-conditioned seat and a heated steering wheel are modeled with experimental results, and the thermal-device models are controlled by the climate control logics. Model validation has been conducted by comparison with the experimental results. In consequence, the results of temperature and thermal sensation in simulation correspond with the experiments. Therefore we can study the effectiveness of prototyped cooperative climate control logics referring to thermal comfort and energy consumption by using the outputs of this MILS before actual vehicle tests.