Three assessment instruments for disturbed children were developed: a 50-item behavior checklist which functioned as a screening device; a 124-item behavior rating scale which provided frequency measures on indices of the teacher's reaction and response to exhibited behaviors; and a behavioral observation form which recorded task-oriented behavior in 10-second intervals for 10-minute periods. The checklist had a split-half reliability of .98 and discriminated between disturbed and non-disturbed children (p=.001). The rating scale reflected treatment differences which were known to exist (p=.01), and had an average inter-rater reliability of .935 for three judges on the behavior of six subjects. Agreement measures between independent observers using the observation form were .90 and above. A treatment model based upon learning theory was developed to modify the behavior of disturbed children in an educational setting. Various response-reinforcement contingencies and reinforcers were used with 11 disturbed boys in grades 4, 5, and 6 and produced measurable change by reducing deviant behavior and increasing time spent engaged in task-orientated behavior. It was not possible to determine which treatment variables produced a given amount of behavior change. The checklist, rating scale, and a classification form are appended. (Author/SN)