This study was conducted during 1982 with the purpose of identifying items for the evaluation of school principals based upon item discrimination power and job situation. A total of 1,393 subjects participated;In the first phase of the study, teachers, principals, and superintendents or principal supervisors rated the performance of a designated principal on 50 items. Discriminating items, or items which measure differences, were identified using a modified ANOVA procedure. A Cronbach alpha reliability coefficient was calculated for all items identified as discriminating;In the second phase of the study, a different group of subjects rated the importance and appropriateness of 10 principal behaviors (identified as discriminating items in the analysis of data from the first questionnaire) in each of three situations to accomplish one of three goals. The situations were designated as very favorable, moderately favorable, and very unfavorable. These data were analyzed using a split-plot with repeated measures design. Duncan's multiple range test was used to locate means which differed significantly when a significant F-ratio had been found in the ANOVA;Forty-nine of 50 items discriminated for the combined elementary and secondary rater group. Different items discriminated among elementary and secondary principals when the data were analyzed separately. Subjects tended to rate a specified principal consistently high or low on all items;Three principal behaviors were judged to be significantly more appropriate and important in reaching the goal of increasing student achievement on standardized tests: (1) takes a strong interest in teachers' professional development; (2) monitors the curriculum and identifies progress toward goals; and (3) promotes activities to solve instructional problems. This research did not provide sufficient evidence to develop criteria for selecting items based on the favorableness of job situation or for reaching the goals of maintaining high teacher morale or student discipline;Discriminating items from this study could be selected for use by local districts. The three items identified as more important for increasing student achievement should perhaps be included on all principal performance evaluation instruments.