The authors used a withdrawal design to evaluate the effects of a modified Color Wheel System (M-CWS) on the on-task behavior of 7 students enrolled in the 4th grade. Standard CWS procedures were modified to include a 4th set of rules designed to set behavioral expectation for cooperative learning activities. Mean data showed that immediately after the M-CWS was applied, on-task behavior increased. In subsequent phases, on-task behavior immediately decreased when the M-CWS was withdrawn and immediately increased when it was reapplied, with no overlapping data across adjacent phases. For each participant, effects sizes calculated across the 3 adjacent phases revealed large effects for 20/21 comparisons. In addition to providing evidence of internal validity, these results demonstrate how educators can adapt the system. (Contains 3 tables and 2 figures.)