The Classroom Behavior Task, a ten-minute classroom simulation developed to measure abilities of potential urban teachers, was tested by comparing performance in the Task with on-the-job performance, using the Classroom Behavior Observation Form and a rating scale divided into "strength and sensitivity" components. Subjects of the study were 20 students enrolled in the Urban Teacher Preparation Program at Syracuse University, who were observed during their participation in the Classroom Behavior Task and during the summer, fall and spring semesters of an urban teaching internship. Significant correlations were found between characteristics of strength and sensitivity in the Classroom Behavior Task and summer classroom performance, with sensitivity characteristics having the stronger relationship. Correlations between the Task and fall or spring teaching were not significant. This corresponded with the results of analysis for linear trend. If the subjects in a preparation program continue to develop in the characteristics of the study, then the relationship with the original measurements must decrease. This study is being replicated, and proposals have been submitted for future research in strength and sensitivity. [Not available in hardcopy due to marginal legibility of original document.] (RT)