Both British and American research into what police actually do has shown that about 70 to 80 percent of police time is spent on service tasks not directly related to crime control. There are grounds for believing, however, that these service tasks aid public cooperation with the police in crime control. Although studies of preventive police patrolling are not unequivocal in their findings, they do cause police managers to question the value of findings, they do cause police managers to question the value of undirected routine patrol. Research on the value of quick response times in answering calls points toward the value of a differentiated response system, in which some calls would require rapid response times while others would permit a more leisurely response without radically affecting case outcomes. The Rand Study of Criminal Investigation shows that 30 percent of arrests were made by uniformed patrol officers and that 50 percent of arrests resulted from the identification of a suspect by the victim or witness. 20 percent of arrests could be loosely attributed to investigative effort and usually resulted from subsequent patrol or administrative actions. These findings have implications for the improved management and coordination of police investigative activity. Future research into police effectiveness should be based in an overall conceptual scheme which will cover the total demands made on police, the nature of police responses, the capacity of police to match responses to demands, and the effectiveness of the results achieved. Notes and references are not listed.