Four of the projects -- Miami, Milwaukee, Seattle, and Washington, D.C. -- were involved in the 3-year evaluation of the project. The evaluation examined the project's impact on the decision to file charges, case processing, and case outcomes. The study used a multiple cohort design, which permitted comparisons between cases processed during a baseline and a program period, and between target cases prosecuted by project attorneys and by other attorneys in the office. The cases compared across time periods were of two types: habitual offender cases that met the project's official criteria and nonhabitual offender cases that did not. Changes over time and between project and nonproject cases were examined using both bivariate and multivariate statistical techniques. The project tended to speed prosecution, reduce plea bargaining, and increase the convictions and sentence severity for the targeted juveniles. Factors in effective programs are identified. 1 table.