A series of supervisory behavior studies that began at Mississippi State University in 1980 are summarized. The studies were instigated because of interest in effective supervision and dissatisfaction with an indepth literature search of supervision literature. The theories disclosed were fragmentary, contradictory, and impractical for use by practicing supervisors; research studies were obscure and not easily generalized to other populations. After years of struggling with conventional theories of motivation and satisfaction, the theories of Frederick Herzberg and Thomas J. Sergiovanni were chosen as most applicable. The paper describes three instruments used in the studies: an ordinal scale, a semantic differential, and the Supervisory Behavior Description Questionnaire (SBDQ), recently developed for use with student teachers and their supervising teachers. The nine studies summarized investigate teachers' preference for certain supervisory behaviors or their perceptions of various supervisory styles. A few studies examine the relationship between perceived supervisory behavior and job satisfaction. Results are mixed, but generally favor a collaborative management style for principals. The SBDQ is praised as an invaluable means of collecting and interpreting data about supervisory behavior. Twenty references are included. (MLH)