The purpose of this paper is to study the basic factors behind educational attitudes by the Q-technique. As an instrument, Fred N. Kerlinger's Q-sort statements were used, somewhat revised to adapt for Japanese teachers. The original 80 item Q-sort had been constructed according to the following paradigm : (A) Attitudes (1) Restrictive-Traditional (2) Permissive-Progressive (B) Areas (a) Teaching-Subject matter-Curriculum (b) Interpersonal Relations (k) Normative-Social (m) Authority-Discipline Thirty-one subjects, consisting of six university professors in Education, six university professors in Psychology, eight elementary school teachers, five junior high school teachers, five senior high school teachers and one layman, were asked to sort 80-statements on an approval-disapproval dimension. Each subject's sort was analyzed with a factorial analysis of variance. As was expected, all the professors' educational attitudes were "permissive" at 1% level, and the other subjects were multifarious in their educational attitudes. About half of the latter had high F-ratios in the permissive direction and the rest had low F-ratios in the permissive or restrictive direction. Twenty-two subjects' sorts were intercorrelated and the resulting correlation matrix was factor-analyzed. Orthogonally rotated factors, I, II and III were named "Progressivism", "Traditionalism" and "Psychology" respectively. A Spearman rank order coefficient of correlation was computed between the attitudes F-ratios and the factors. The rhos were .86, -.24 and .44 for I, II and III respectively.