The major purpose of this study was to examine the relationship between a systematic, media facilitated, approach for training supervisors to conduct postobservation conferences (Coaching and Counseling for Conference Improvement) and conference effectiveness. A non-equivalent pretest-posttest control group design was used. The data for this study were gathered from twenty-four teacher supervisors in North Iowa during the 1983-84 school year. Six instruments were used to collect pretest and posttest data;The Coaching and Counseling for Conference Improvement (CCCI) training differed from the "one shot" inservice of the control group in that CCCI supervisors videotaped practice conferences, received critical feed back from peers, and set goals for improvement. Pretest and posttest data, including ratings of the videotaped conference performance of experimental group supervisors, were analyzed using t-tests;The study yielded four important findings: (1) Trained raters rated the skill level in conducting conferences of the experimental group supervisors significantly higher following CCCI training. (2) Supervisors trained via CCCI were significantly more self confident in their ability to conduct conferences than control group supervisors following training. (3) The skill in conducting conferences of the CCCI supervisors was rated significantly higher by teachers than was the skill of control group supervisors;No significant changes were found in conference climate of either the experimental and control groups. Experimental and control group teachers did not significantly differ on their perceptions of the potential for the conference to improve their teaching performance.