This paper applied the SAS STATESPACE models to forecast graduate student increase in Taiwan. Addressing the following research questions, the author tried to develop some quantitative strategies for graduate education systems in future: 1. How to set an acceptable limit on the increase in graduate students? 2. What has been the influence of government policies concerning graduate study abroad? 3. How to balance the number of graduate students of the arts, the social sciences and that of the applied sciences? 4. How to set a reasonable graduate student ratio in both private and public universities? Using the data of Educational Statistics of the Republic of China published by the Ministry of Education in 1992, the author forecast the number of graduate schools, the number of graduate students, and the ratio of graduate students compared with the total population, major areas adifferent sectors of the universities from 1992 through 2000. The paper also applied the ARIMA model to forecast the number of graduate students and compared with that of STATESPACE. The STATESPACE model may help policy makers to formulate the general graduate education policy.