Except for a few studies, most research investigating correspondence training procedures has been more analogue in nature. The purpose of the present set of studies was to examine whether a "say-do" correspondence training technique could be used with children in special education classes to improve classroom behavior. The specific behaviors targeted for change included: out-of-seat behavior (Experiment 1), sitting posture (Experiment 2), and on-task behavior (Experiment 3). The say-do procedure used in Experiment 1 resembled that of previous studies, whereas that in Experiment 2 was more elaborate in the specificity of verbal statements required from the children and the feedback given them. The training procedure in Experiment 3 used a format similar to the say-do approach, but stressed visual rather than verbal cuing because it was used with nonverbal children. All three studies used single-subject designs and examined maintenance and/or generalization questions. Experiments 2 and 3 also evaluated whether concomitant changes in performance on academic tasks occurred. The results of the three studies provide strong evidence that correspondence training can be effectively used with educationally handicapped children. Moreover, the successful modification of the "say-do" to a "show-do" procedure in Experiment 3 points out the flexibility of the correspondence training approach.