The effect of contingent nonverbal teacher approval on student attentive behavior was examined in a dassroom with 12 retarded children. After baseline data were gathered on contingent verbal and nonverbal teacher approval and student attentive behavior, the teacher was instructed to increase her use of contingent nonverbal approval (smiles and physical contact) and to maintain her baseline level of verbal approval. After a reversal phase, the nonverbal approval phase was reinstated. Nonverbal teacher behaviors increased during the experimental phases, whereas verbal teacher approval (alone or in conjunction with nonverbal behaviors) did not increase. Attentive behavior increased for 11 of 12 students during the phases in which contingent nonverbal teacher approval increased. Correlational data suggested that nonverbal teacher approval accounted for behavior change of the students to a greater extent than did changes in the amount of teacher approval per se or in the teacher's use of verbal approval.Teacher approval or attention has been used extensively to alter student behavior in the classroom. In many studies, teacher approval is operationalized as verbal praise (Hall, Panyan, Rabon, and Broden, 1968;McAllister, Stachowiak, Baer, and Conderman, 1969;Ward and Baker, 1968;Wasik, Senn, Welch, and Cooper, 1969). Yet, in other studies, verbal approval is supplemented with nonverbal attending behaviors on the part of the teacher such as smiling, patting, holding, touching, moving toward the child, making eye contact, varying tone of voice, among others (Becker, Madsen, Arnold, and Thomas, 1967;Buell, Stoddard, Harris, and Baer, 1968;Cooper, Thomson, and Baer, 1970 their individual effect on student behavior is rarely evaluated. Even when observational codes include nonverbal teacher behaviors such as facial expressions and physical contact (Becker et al., 1967) or proximity of the teacher to the pupil , it remains unclear whether these behaviors contribute to behavior change in the students independently of verbal approval alone. One reason that nonverbal attending behavior usually is not evaluated is that verbal approval, as opposed to facial expressions and physical contact, constitutes most (about 85 %) of teacher attending behavior (Madsen, Becker, and Thomas, 1968). Moreover, teacher training is more likely to emphasize and alter verbal behavior as opposed to nonverbal behaviors (facial expressions and physical contact) (Kazdin, in press). Nevertheless, nonverbal teacher behaviors alone may provide an important source of reinforcement in the classroom. Indeed, several texts and manuals on behavior modification in the classroom advocate the use of nonverbal approval behaviors either to supplement praise or to provide reinforcement in their own right (