The four programs, selected through a detailed screening and selection process, were located in Orleans Parish, La.; San Diego County, Calif.; Franklin County, Ohio; and Kalamazoo County, Mich. The report examines the development of the program concept and the assumptions underlying program effectiveness, as well as the program processes themselves. These include the routine prosecutor practices which form the context for program implementation, the targeted practices that were instituted for the program, and the target populations. The report also analyzes the extra-program processes in law enforcement and corrections which may aid or impinge upon the ability of the program to achieve its objectives, and program effects on the performance of the criminal justice system. Major findings suggested that disposition rates were not significantly affected by the Career Criminal Program in any of the four sites, but that the program affected the strength of offender convictions in two sites. In some cases, these convictions were accompanied by longer sentence lengths. Increases were also observed in rates of conviction to the most serious charge and plea to the most serious charge in two sites. No increases in incarceration rates were observed. However, in three of the four sites, high preprogram rates of incarceration (90 percent and above) for convicted career criminals made program effects unlikely. In the fourth site, prison overcrowding led to a decline in incarcerations for noncareer criminals; in this context, the career criminal incarceration rates appeared stable. Increased commitments to State prison were observed in two sites. Finally, one site showed marked improvements in processing time for career criminal cases, while no significant differences were observed in the other three sites. The evaluation had some limitations. The research design involved only a small number of sites, and since the evaluation was designed to examine the impact of local program activities as they were implemented in their entirety, it does not specify which strategies were important to program outcomes. Footnotes, 17 tables, and 6 figures are included. (Author abstract modified)