James Q. Wilson's findings in 'Varieties of Police Behavior' provided the basis for analysis of departmental records, policies, and practices of the three agencies. Data were collected and analyzed, then compared with Mintzberg's and Wilson's studies. Literature concerning the police chief executive, written since 1900, and current change-management literature were reviewed. The literature indicates that recommended standards of performance for the police chief executive often are not met. Appointments of qualified personnel are inhibited by lack of understanding of difficult position requirements and by local political consideration. For these and other reasons, quality change management is rare in police agencies. The work profiles of the police chiefs studied were similar to those of executives studied by Mintzberg. The police chiefs spent more time managing improvements, devoted more time to short-range issues, and handled many more documents, however, than those studied by Mintzberg. Much activity was external to their agency. Stated agency policies indicated that all three chiefs subscribed to Wilson's service orientation. Some arrest data, however, indicated that Wilson's legalistic orientation was emphasized in practice. Thus, a conflict was found between policies and practices. Coherent philosophy was lacking in all three agencies studied, which contributed to conflicts between policies and practices, inhibiting improvement. Appendixes contain survey instruments and background documents, and 111 bibliographic listings are provided.