Leadership has been studied as a crucial factor affecting the performance of an organization. Traditionally there has been considerable interest in the role of leadership in educational administration. Many a study has been made they have however, primarily borne the characteristic of survey studies or prescriptive studies. The lack of scientific and theoretical studies has been conspicuous, naturally provoking increasing need for theoretical studies of leadership. The recent research on leadership has been activated under the name of the cotingency theory. The contingency theory does not seek all-purpose leadership applicable under any circumstances, but rather searches for differential functions of an organization and leadership styles, which are related to the nature of the organization. A typical one of the contingency theories is Fiedler's leadership theory: that is, the Fiedler's Contingency Model. Fiedler presented a theory of leadership effectiveness which takes account of the leader's personality as well as the situational factors in the leadership situation. He emphasizes that the same type of leadership style or leadership behavior will not be suitable for all situations. This theory attempts to specify in more precise terms the conditions under which one leadership style or another will be more conducive to group effectiveness. The theory thus reconciles after conflicting claims and results which would favor one style of leadership over another. It may be briefly summarized as follows; 1. There are two different kinds of leadership styles which are measured by the Least Preferred Co-worker (LPC) scale. (1) Relationship-motivated (High LPC, score of 73 or above) leaders tend to be most concerned with maintaining good interpersonal relations, sometimes even to the point of letting the task suffer. In relaxed and well-controlled situations persons of this type tend to reverse their behavior and becomes more task consciout. (2) Task-motivated (Low LPC, score of 63 or below) leaders place primary emphasis on task performance. These leaders are the nononsense people who tend to work best from guidelines and specific directions. If these are lacking, their first priority is to organize and create these guidelines and then assign the various duties to their subordinates. However, under relaxed and well-controlled situations, task-motivated leaders take the time to be pleasant and pay more attention to the morale of their employees. Leaders whose score fails between 64 and 72 will have to determine for themselves which category they most nearly resemble. 2. There are three kinds of leadership situations: (1) High control situations allow the leaders a great deal of control and influence and a predictable environment in which to direct the work of others. (2) Moderate control situations present the leaders with mixed problems-either good relations with subordinates but an unstructured task and low position power, or eversely, poor relations with group members but a structured task and high position power. (3) Low control situations offer the leaders relatively low control and influence, that is where the group does not support the leaders, and neither the task nor his position power give them much influence. Stress of high group conflict may also ocntribute to low control. 3. There are three dimensions that determine the situational control of a job. These are: (1) Leader-member relations measure how well the group and the leader get along. (2) Task structure measures how clearly the procedures, goals, and evaluation of the job are defined. (3) Position power measures how much authority to hire and fire and discipline the leader has. 4. In matching leadership styles to appropriate situations we find that: (1) Relationship-motivated leaders perform best in moderate control situations. (2) Task-motivated leaders perform best in high and low control situations. Cept for the unusual cases, it is simply not meaningful to speak of an effective leader or of an ineffective leader; we can only speak of a leoder who tends to be effective in one situation and ineffective in another. If we wish to increase organizational and group effectiveness we must learn not only how to train leaders more effectively but also how to build an organizational environment in which the leader can perform well. Leadership training is possible to change or modify your leadership situation if you find that your leadership style does not match the situation in which you are working. You can engineer your job by adjusting the three dimensions of situational control and making it higher or lower, thereby matching your leadership style. Transfer and rotation, selection and placement, are management tools to improve the performance of your subordinate leaders, thereby increasing organizational effectiveness. Even though Contingency Model is not an all-purpose solution to all kinds a leaders-hip problems, Fiedler's theory has achived a prominent effect. His Contingency Model is an epochrnaking approach since it puts group task situation in between the independent variables of leadership and the end-result variables of group performance, and it has brought about a number of experimental and positive data. The theory of course has some defects. For example, his LPC scale has a theoretically weak relevance, neglecting a primary factor in long-ranges situation affecting a leader and his group depending upon the use of uni-dimensional scale in determining the leadership style. Several scholars have attempted to solve these problems by applying and developing the Contingency Model. To cite a few examples: in Saha's study(1972) , "an effort was also made to extend the model by investigating the effect of three new variables , namely intelligence, ability and motivation as determinants of situational favorability for a leader ; in Bons' study (1974 ), "the effect of changes in leadership environment on the behavior of relationship-motivated and task-motivated leaders," and Kennedy's (1982), "middle LPC leaders and the Contingency Model of leadership effectiveness" , etc. It is possible to generalize Fiedler's theory that the effectiveness of group performance is affected by the change of leadership typology and group situation. It should be accompanied by the statistical tests. It remains to be studied further what factors affect the accuracy of model, which factors directly change the performance of model and so forth. When such defects as above-mentioned are removed and its generality is proven with the help of the fruitful study hereafter his theory should be applied to the educational administrations, training, rotation, and transfer as well as to the empirical analysis of the leadership effectiveness of school principalship.