Hospital staff members are responsible for a heavy workload in an atmosphere of utmost urgency. The health maintenance of these employees can highly elevate their administrative efficiency and the quality of medical care they provide in hospitals. The aim of this paper is to provide a method for strengthening the employee health management system with genetic algorithms and a back propagation neural network. The management system can assist hospital administrators to formulate a health policy and serve as an important reference for health self-management for hospital staff. Based on a survey of 2570 hospital staff members from four hospitals, this paper considers their lifestyles, determined by questionnaires, as conditional attributes and the corresponding results of physical examination as decision attributes. Genetic algorithms are applied to predict the relationship between abnormal ratios of health examination and lifestyles. Furthermore, the back-propagation network is also applied to predict not only morbidity but also tendencies in personal health. The 3-D Bezier Surface is presented in the health management system to visualize the relationship between lifestyle and disease. This visualized prediction system could function as an appropriate index of self-health management and self-improvement in life quality for hospital staff.