In a car equipped with an ECC (Electronic Climate Control) a correct feedback of the compartment temperature is crucial for the performance of the controller. In this thesis work a model of the compartment temperature has been developed to overcome the problem with uncertainties in the measurement signal from the compartment temperature sensor. The model can be divided into two major submodels where the first submodel contains several state-space systems for different ambient temperatures. This part also uses look-up tables and piecewise linear interpolation to ensure a smooth transition between different working points. The second submodel contains a system to detect and compensate for open doors or windows. The state-space systems are developed in the System Identification Toolbox in Matlab, where data from earlier measurements in test vehicles work as the input to the identification algorithm. The model is tested on data in different climates and the output is compared to measurements from thermocouples in test vehicle for model validation. The compartment temperature sensor value is compared to the model output to point out the improvements in the compartment temperature feedback. Results show that the model output usually deviates within 1 °C from the thermocouple value compared to the sensor's 2 °C to 3 °C. The model is developed in both Matlab / Simulink and Xmath / SystemBuild environment.