Cave ventilation significantly increases the depth of natural thermal oscillations and decreases the time of propagation compared to heat conduction in the rock mass. This makes it necessary to develop and test thermal models for the prediction of temperature fields in ventilated karst massifs. Here, we develop a thermal model of a single conduit ventilated by chimney effect. The model is based on the diffusion equation in the rock mass coupled to the conservation of energy and water vapor mass in the airflow. The effect of the latent heat of evaporation and condensation is considered. In parallel, the main conduit of a ventilated cave has been equipped with a flowmeter and several temperature sensors. The model is tested against field data collected during a complete year. The relevance of the model assumptions (geometry simplification, initial and boundary conditions, use of transfer coefficients to couple the air and the conduit wall) is thoroughly analyzed. The model correctly predicts the temperature fluctuations at daily and yearly scale, but underestimates the annual mean temperatures inside the cave. A biased assessment of the ground temperature seems to explain this discrepancy. The effect of condensation and evaporation on the cave climate turns out to be low on cave temperature, but significant on air humidity with consequences for ecology or paleoclimatology. This study is a first step towards the elaboration and validation of models providing a quantitative assessment of caves’ thermal response at any location and time scale.