In an attempt to determine the effectiveness of a pretreatment assessment procedure known as the scatter plot (Touchette, MacDonald, & Langer, 1985), direct observational data was collected by 13 middle school teachers on four "problem" students. After four weeks of data collection, interobserver agreement probes were calculated and a visual analysis of the plotted data was performed to ascertain a possible pattern of problem behavior. Additionally, in an attempt to assess the teachers' perceptions of the scatter plot, the 13 teachers were asked to complete a questionnaire. Although a visual analysis of the plotted data suggested a possible pattern of problem behavior, interobserver agreement probes failed to achieve a desired overall reliability of 90% or higher. Despite a low IOA, results of the questionnaire administered to the 13 teachers generally supported the use of the scatter plot as a means of assessing student behavior. Possible reasons for failing to attain an IOA of 90% or higher include the total number of students in a class, the number of subjects observed per period, the teacher's location in the classroom, and the subjects ability to recognize if the teacher was "looking." Recommendations are provided regarding future research concerning the scatter plot and other more formal approaches to assessing student behavior.