The purpose of this research was to develop and validate a self-report, group-administered, non-verbal inventory to measure Self Perception In School (SPS) among primary grade children. Inventory items were based on Sarbin's Role Theory. Since role may be studied in terms of the actions expected of an occupant of that position, the role of student was defined as those behaviors teachers expect of students. From this universe, a set of items was devised. Three pilot studies were conducted. The final version of the inventory uses 19 cartoon-like drawings in transparency form (dichotomous presentation and responses) to probe an equal number of behaviors. In the major study, the Self Perception in School (SPS) inventory was administered to a sample of 656 first, second, and third graders in two schools of a white middle-class suburban system. It was concluded that: (1) the SPS inventory is a reliable, valid, inexpensive, non-verbal measure of the construct at the primary level; (2) the structure of the "self" is not unidimensional; (3) the SPS is significantly related, at the primary level, to teachers' ratings of children's behavior and to reading and arithmetic achievement; (4) girls have a more positive self perception in school than boys at the primary level; and (5) while the mean SPS of girls is relatively similar across grades, the mean SPS of boys consistently decreases from grade one through grade three. Means, standard deviations, and reliabilities are listed. See TM 000 409 for a report on correlates of the SPS. (Author/DG)