This evaluation of the second-year (1986-87) implementation of the New York City (New York) Dropout Prevention Initiative (DPI) in 13 high schools focuses on student characteristics and outcomes, with a discussion of a set of statistical data on program services, findings on school-level indicators of program effectiveness, and an analysis of program costs. The DPI has been organized to test the effectiveness of the following models of service delivery: (1) pure systemic, in which services are directed toward changing the entire school environment; (2) modified systemic, in which some case management services for at-risk students are provided by community-based organizations (CBOs); (3) modified CBO, in which case management services provided by CBOs are augmented with some funding for systemic services; and (4) pure CBO, in which the program consists entirely of case management services provided by CBOs. Data were analyzed from the following sources: (1) site visits; (2) interviews with school personnel; (3) logs of services delivered to students; (4) a student attitude survey; (5) the New York City Board of Education's student data base; and (6) monitoring reports on CBOs. Overall, the DPI was not successful in improving student attendance or academic achievement enough to ensure that students would be able to earn a high school diploma within a reasonable length of time, although targeted students were less likely to drop out than eligible-but-not-targeted students. Recommendations for improvement are included. Statistical data are included on 36 tables and 15 graphs. (FMW)