A flow network method was developed to predict the underhood temperature distribution of an automobile. The method involves the solution of simplified energy and momentum equations of the air flow in control volumes defined by subdividing the air space between the surfaces of the underhood components and the front-end geometry. The control volumes are interconnected by ducts with branches and bends to form a flow network. Conservation of mass and momentum with appropriate pressure-loss coefficients leads to a system of algebraic equations to be solved for the flow rates through each volume. The computed flow rates are transferred to a thermal model to calculate the temperatures of the air and the major vehicle components that affect the underhood environment.The method was applied to a 1986 3.0L Taurus and compared with vehicle experiments conducted in a windtunnel. Computed results agree with vehicle data to within experimental uncertainty for both transient and steady state operations. Applications illustrated in the paper include the selection of radiator and its effects on air-conditioning, the tradeoff of body styling on underhood temperature and the requirement of high fan speed to compensate for air blockage.