Project New Pride, which has been designated a LEAA Exemplary Project, illustrates the need to adapt the evaluation process in order to provide a useful evaluation. The project's original evaluation plans called for an experimental design in which a random sample of juveniles would be selected for treatment within the project, while a control group would receive the traditional court services. Modification of this plan was necessary because of the small size of the eligible pool of juveniles. In addition, referral sources were not interested in random assignments, but in making the most appropriate referral for any given client. A factorial evaluation design was rejected due to similar problems. The evaluator then shifted the focus to project implementation, using a descriptive rather than comparative approach. Evaluation questions thus concerned the information used to determine treatment plans, the extent to which treatment plans were implemented, and the differential effects of project activities across types of juvenile offenders. These questions required the generation of timely, uniform, and complete data sets. A computerized system of report generation was needed. As a result, a management information system was designed to track and analyze client outcome data, support internal project management needs, monitor service delivery activities for each client, and generate timely statistics. The management information system was closely linked to the evaluation questions. It is concluded that the descriptive methodology was useful to project management and staff. The management information system will be operating on a full data base by September, 1980; from early indications, it will be an effective diagnostic tool for project development and change.