Abstract: Based on the national statistical yearbook data from 2001 to 2016,this article studied the spatial pattern changes of the distribution of high-educated talents in China and its spatial overflow effect as well as the influence mechanism by applying Durbin model, Gini index, CV index and spatial auto-correlation. The result shows that:（1）Distribution status of China's high-educated talents showed that talents are agglomerate in the southeast but sparse in the northwest. From 2000 to 2015,the area of sparse area has decreased while he area of agglomerate area has increased；（2）The distribution of high-educated talents has some spatial agglomeration characteristics. In 2000,Hot spots are the Beijing-Tianjin-Hebei and Yangtze river delta urban agglomeration. However, only Beijing Tianjin and Shanghai are the highest level of agglomeration，other provinces are all cold spots. What’s more, from 2000 to 2015, the area of cold spots and hot spots both have increased which implies that talent still has a tendency to centralize in some regions；（3）Wages, the development of colleges and universities, and public service level are the main factors influencing the spatial distribution of high-educated talents and no evident can prove other factors have significant influence on the distribution of talents..The distribution of high-educated talents has significant spatial overflow effect while significant factors including Wages, the development of colleges and universities, and public service level also have significant spatial overflow effect.