Developing tourism is one of the most important policies of the Government. But in order to understand the management effects and to evaluate the accomplishment of management, we should control the variation of tourists exactly. Kenting National Park is the first National Park in Taiwan. Owing to possess abundant biological resource and high-diversity ecosystem, it became a famous tourist attraction. While the Government is getting 'Double-tourist plan' in moving, the National Park is as important as can be. But the administration office must in view of ”the purpose can be approached or not?” so it must be evaluated in advance. For this purpose, we discussed the variation of tourists in first. Then we used the 'grey prediction model' and 'grey relation method' to forecast the quantities of tourists, to evaluate the influential factors of tourists' quantities. So this research will be useful in the management of Kenting National Park.