A 15-beat 32-square mile area in Kansas City was divided into three groups--reactive, proactive, and control--of five beats each. In the five reactive beats, routine preventive patrol was eliminated. In the proactive beats, the department doubled or tripled the normal level of police patrol vehicles assigned to these beats by the allocation of additional patrol cars and the presence of units available for assistance calls in the reactive beats. In the remaining five beats, the control group, the normal level of preventive patrol, one car per beat, was kept. Results indicated no statistically significant differences in crime in any of the 69 comparisons that were made between reactive, contorl, and proactive beats. Further, police arrests showed no statistically significant differences in the 27 comparisons made between reactive, control, and proactive beats. In addition, citizen fear of crime was not significantly affected by changes in the level of routine, preventive patrol. The Kansas City Police Department now uses a Direct Patrol Deployment System which involves a highly efficient use of manpower and crime statistics, crime prevention techniques, and citizen involvement. Citizen needs for routine, preventive patrol are still met because the criminal element is not aware of variations in patrol. Additional research is necessary to determine the effectiveness of different law enforcement procedures. Further, the Kansas City experiment demonstrates that such research can be conducted within police departments. A chart is included.