A study was conducted to develop a shortened version of the psychological stress level scale for junior high school students in physical education classes introduced by Sasaki (1997), to examine its reliability and validity, and furthermore, to establish its evaluation standard. First, stressor descriptions in physical education classes were collected from a factor analysis of data from 756 junior high school students and a shortened version was made, consisting of 5 subcategorized scales: "lack of a feeling of achievement," "teacher attitudes," "reception of slander," "insincerity of classmates," and "deficiency of physical condition." Each of these scales included 18 items. Next, the reliability of the new version was examined through Cronbach's α, the split-half method and the test-retest method. The resulting coefficients of the 5 scales were .78-.81, .74-.83, and .72-.87, respectively. Thus, this version's reliability was verified as satisfactory. Furthermore, the version's cross-validity was examined based on multiple regression analysis where the scores on this version's scales were treated as independent variables and the scores for psychological stress responses in physical education classes as criterion variables. The resulting validity coefficient showed that this version's validity was satisfactory. Finally, from analysis of the score distributions, an evaluation standard of three phases was established, which can be applied similarly to all students in three grades but differently to male and female students.