DATA FROM STUDIES AT THREE INSTITUTIONS SUPPORT THE HYPOTHESIS THAT FACULTY MOBILITY IS CHIEFLY BETWEEN INSTITUTIONS OF APPROXIMATELY THE SAME SIZE AND FUNCTION, AND NOT FROM SMALLER, LESS PRESTIGIOUS INSTITUTIONS TO LARGER, MORE RESPECTED INSTITUTIONS. THIS FLOW IS REGARDED AS AN OUTCOME OF THE FOLLOWING FACTORS--(1) INSTITUTIONAL AFFINITY, (2) INFORMAL COMMUNICATION AMONG STAFFS OF INSTITUTIONS OF SIMILAR TYPE, (3) SIMILARITY IN ROLE DEMANDS (PARTICULARLY TEACHING SPECIALTY), AND (4) THE CONDITIONING EFFECT OF PAY, BOTH IN EXCLUDING HIGHER-PRICED AND LOWER-PRICED CANDIDATES FROM CONSIDERATION AND IN FAVORABLY EVALUATING CANDIDATES. RETENTION OF AN INDIVIDUAL IN AN INSTITUTION IS A FUNCTION OF THE COMPATIBILITY BETWEEN INDIVIDUAL CONDITIONING AND EXPECTATIONS, AND THE CHARACTERISTICS OF THE FACULTY SOCIAL SYSTEM OF THE INSTITUTION. THREE MAJOR SUGGESTIONS ARE--(1) VACANCY AND AVAILABILITY INFORMATION SHOULD BE CIRCULATED NOT ONLY WITHIN DICIPLINES AND GEOGRAPHIC REGIONS, BUT ALSO AMONG INSTITUTIONS OF SIMILAR TYPE, (2) CAREER DEVELOPMENT OF COLLEGE TEACHERS, PARTICULARLY IN ITS INITIAL ORIENTATION IN GRADUATE SCHOOL, SHOULD BE DIRECTED TOWARD THE LIKELY FUTURE INSTITUTIONAL SETTINGS, AS WELL AS THE PROFESSIONAL FIELD, AND (3) MANPOWER PLANNING FOR EDUCATION MAY BE ORIENTED TOWARD A MORE SPECIFIC "DEMAND" BASE. THIS PAPER WAS PRESENTED AT A MEETING OF THE SOUTHERN SOCIOLOGICAL SOCIETY (NEW ORLEANS, APRIL 7-9, 1966) AND WAS REPRINTED FROM "COLLEGE AND UNIVERSITY," FALL 1966. (LC)