The purpose of this study was to examine the extent to which the structure or patterning of supervisor and supervisee verbal interaction during supervision is enhanced or moderated by: (1) the problem severity (degree of impairment and level of distress of the client); (2) the anxiety of the supervisee and supervisor toward the client and toward supervision (i.e., toward each other); (3) self-ratings of competence by the supervisee and supervisor; (4) amount of supervision provided to the supervisee; and (5) supervisors' years of supervision experience. Structure (i.e., communication and relationship patterning) was operationalized as the degree of "non-randomness" (absence of entropy) in the verbal responding of the supervisor and supervisee. Forty-four actual counseling supervision dyads were analyzed through transcriptions of taped supervisory sessions. Multivariate procedures were used to examine factors contributing to the structure of supervisory sessions and the structuring influences of the supervisor and supervisee. Results suggest that supervisor-supervisee interaction may be less structured than generally thought, and that anxiety, skill, and experience variables play a negligible role in influencing the amount of structure that does occur. Supervisor sex was found to be a reliable predictor of whether the supervisor or supervisee had the greater influence on the supervision structure. Contains 36 references. Five tables present statistical analysis results. (Author/JBJ)