The study provides an overview and analysis of the different approaches, taken by the central government of Taiwan, the U.S and England, used to evaluate local government administrative performance in special education. It was conducted through literature review. The results show that Taiwan has done evaluations in 2002, 2004, and 2006, and indicated nine advantages or specialties: (1) The evaluation served the functions of “understanding,” “reviewing,” “improving,” “elevating,” and “emulating.” (2) The evaluations conducted by The Society of Special Education of R.O.C. were objective and trustworthy. (3) Evaluation committees were acknowledged to be professional, proficient and objective. (4) Evaluation items were decided by CIPP and countenance models. Both idealism and reality were taken into consideration. Items with high discriminant values were selected, alleviating the burden on local governments. (5) Indices for all item scores were clearly quantified or qualified. (6) The evaluation progress, the process of paper work, and the posting of information on the internet were all appropriate. (7) The outcomes of the evaluation were well presented, covering not only the ranks, but also the statements of improvement. (8) Evaluation outcomes were simplified to include four levels for human resources in specified fields, and two levels (pass/fail) for appointed and self-chosen fields. (9) The evaluation not only offered sufficient rewards to the responsible faculty, but also fully integrated with funding system. With expectations to improve the current system, the methods of the U.S. and England serve as references to provide two propositions: establish better routine review mechanisms and set up a routine alarm mechanism.