The research included two national surveys of U.S. police departments in 1971 and 1974 and a series of seven case studies in different police departments around the country. Since the mid-1960's, when the first police computer system was installed in the St. Louis Police Department, the computer has become a permanent part of law enforcement technology. However, successful implementation has often been limited to routine areas, and results of computer applications to nonroutine areas have been disappointing. Routine applications involve the relatively straightforward, repetitive manipulation and inquiry of prescribed data, often by means of a definite procedure. The same manipulation was usually done by hand before the computer was used. Such routine applications include traffic files, crime statistical files, police administration, and miscellaneous operations. However, in nonroutine applications, the computer becomes a tool for decisionmaking, strategic planning, and person-technology interaction. Nonroutine application areas in law enforcement include resource allocation, investigation of crime, and command and control -- including computer-aided dispatch and automatic vehicle monitoring. Routine and nonroutine applications should be viewed as delineating two ends of a continuum. As applications move toward the nonroutine end of the continuum, systems design becomes more intricate, and behavioral, personality, and organizational considerations become more significant. This paper observes that the 1967 Crime Commission Report raised expectations about the extent of computer applications to law enforcement too high and failed to focus adequate attention on the implementation and diffusion of innovation throughout the law enforcement community. The eventual influence and impact of technology in policing will not come from technology per se, but from an interaction between police work, the nature of a particular department, and any specific innovation. Moreover, the use of technology may have an important influence on power and prominence within organizations. Finally, the diffusion of police computer technology in the future will involve four basic steps: inventing (the creating of ideas, technologies, and models), informing (publicizing the technology and educating the law enforcement agency), and integrating (the overall social and economic acceptance and adjustment to the innovation by the agency). Figures and 26 notes are provided.