Live (Direct) Supervision has always been key to training therapists. However, little research has been done to determine which type of direct supervision is most effective in changing therapist behavior. This study compared bug-in-the-ear (BITE), phone-in, and computer assisted (CA) supervision. The purpose of this study was to (a) complete functional analysis of each supervision type via an alternating feedback session, to determine each therapists' most effective form of supervision; (b) determine the effect of using the most-effective form of supervision on the therapists' adherence to treatment protocols during follow-up sessions; (c) determine the effect of therapist adherence on client in-session behaviors, and outcome assessments. Results of the functional analysis indicated that phone-in supervisory feedback was the least effective form of supervision. BITE and CA supervision were both found to be effective for at least one therapist. In follow-up sessions BITE supervision did not create sustained changes in overall adherence. CA feedback appeared to maintain overall therapist adherence. In affecting changes to client behavior and outcome scores BITE supervision gave mixed results. CA supervision created desired changes in client behaviors and outcome assessments. For practitioners in the field, this study demonstrates that client behavior does not improve without therapist intervention. Even when client behavior in session appears to be improving, marital distress may increase. This is especially true when therapist adherence does not improve. This study further demonstrates that when the therapists did not intervene appropriately the clients either worsened, or at best, stayed the same. When the therapists made even small improvements client behavior improved. Improving adherence to treatment protocols will always serve the best interest of the client.