In the motorized spindle system of a computer numerical control (CNC) machine tool, internal heat sources are formed during high-speed rotation; these cause thermal errors and affect the machining accuracy. To address this problem, in this study, a thermal resistance network model of the motorized spindle system is established based on the heat transfer theory. The heat balance equations of the critical thermal nodes are established according to this model with Kirchhoff’s law. Then, they are solved using the Newmark-β method to obtain the temperature of each main component, and steady thermal analysis and transient thermal analysis of the motorized spindle system are performed. In order to obtain accurate thermal characteristics of the spindle system, the thermal-conduction resistance of each component and the thermal-convection resistance between the cooling system and the components of the spindle system are accurately obtained considering the effect of the heat exchanger on the temperature of the coolant in the cooling system. Simultaneously, high-precision magnetic temperature sensors are used to detect the temperature variation of the spindle in the CNC machining center at different rotational speeds. The experimental results demonstrate that the thermal resistance network model can predict the temperature field distribution in the spindle system with reasonable accuracy. In addition, the influences of the rotational speed and cooling conditions on the temperature increase of the main components of the spindle system are analyzed. Finally, a few recommendations are provided to improve the thermal performance of the spindle system under different operational conditions.