This paper presents a new theoretical model to determine the optimal axial preload of a spindle system, for challenging the traditional method which relies heavily on experience of engineers. The axial preloading stiffness was treated as the sum of the spindle modal stiffness and the framework elastic stiffness, based on a novel concept that magnitude of preloads can be controlled by measuring the resonant frequency of a spindle system. By employing an example of a certain type of aircraft simulating rotary table, the modal stiffness was measured on the Agilent 35670A Dynamic Signal Analyzer by experimental modal analysis. The equivalent elastic stiffness was simulated by both finite element analysis in ANSYS? and a curve fitting in MATLAB?. Results showed that the static preloading stiffness of the spindle was 7.2125×107 N/m, and that the optimal preloading force was 120.0848 N. Practical application proved the feasibility of our method. This paper presents a new theoretical model to determine the optimal axial preload of a spindle system, for challenging the traditional method which relies heavily on experience of engineers. The axial preloading stiffness was treated as the sum of the spindle modal stiffness and the framework elastic stiffness , based on a novel concept that magnitude of preloads can be controlled by measuring the resonant frequency of a spindle system. By employing an example of a certain type of aircraft simulating rotary table, the modal stiffness was measured on the Agilent 35670A Dynamic Signal Analyzer by experimental modal analysis. The equivalent elastic stiffness was simulated by both finite element analysis in ANSYS® and a curve fitting in MATLAB®. Results showed that the static preloading stiffness of the spindle was 7.2125 × 107 N / m, and that the optimal preloading force was 120.0848 N. Practical application proved the feasibility of our method.