The guidance counselor is frequently seeking information from the classroom teacher about the overt behavior of a child in the classroom. In this study elementary school teachers were asked to rate their students on items describing specific observable classroom behaviors in two sessions with a two-week interval between ratings. The items used to rate students were determined in a pilot phase when elementary teachers were asked what concepts they considered important and not important for the satisfactory behavior of a child in the classroom. The results strongly indicated that teachers were not stable in rating the overt behaviors of pupils. The item reliabilities tended to increase as the number of rating categories available to the teachers increased from five to seven to nine; however, no statistical differences was found. Assuming that teachers do rate and possibly refer children to an elementary counselor based on a single episode, it would seem imperative that the elementary counselor determine as quickly as possible the generality of the behavior problem. (Author)