Vehicle development teams find it challenging to predict what their Heating, Ventilation and Air-Conditioning (HVAC) module performance will be for cold ambient (∼ -20 deg. C) test cycles such as defrost and cabin warm-up before the car is built. This uncertainty in predictions comes from varying engine heat rejection to coolant due to cold cylinder wall temperatures, calibration changes and degraded performance of various components within the cooling system such as the coolant pump owing to higher viscosity of the coolant. Measuring engine heat rejection at cold ambient is extremely difficult as the engine warms up as soon as it is fired. Multiple measurement points require long lead time to soak to the cold target temperature. It is a common practice to adjust engine calibration parameters to warm up coolant as fast as possible for an adequate defrost and cabin warm-up performance. A computational model can really help quantify the effect of each of the key parameters to enable the HVAC engineer to achieve the target performance most efficiently.In view of this, a novel approach is proposed to simulate and improve cold ambient warm-up predictions using one-dimensional (1D) computational models. Initially an engine thermodynamic model was developed and validated for a 2L gasoline turbocharged direct injection engine using commercial code Ricardo WAVE®. This model was used to predict cold ambient engine heat rejection and quantify the effect of each of the calibration parameters such as spark timing, air-fuel ratio, manifold pressure and cylinder wall temperature. The modified engine heat rejection was then applied to a cooling system, developed in Flowmaster® where the effects of the engine and coolant thermal inertia and component performance degradation were taken into account to quantify the final effect of various key parameters on cabin air temperature.It was observed that proposed approach would reduce testing cost by guiding the testing process into targeted tests to achieve desired cold warm-up capability efficiently.