The effects on selected student outcomes are evaluated after 2 years of operation of Project RAISE, a multifaceted approach to helping at-risk students featuring outside adults as school-based advocates and one-on-one mentors at seven middle schools in Baltimore (Maryland). Seven community sponsoring groups (churches, universities, businesses, and a social fraternity) each sponsor approximately 60 at-risk youth. Positive effects are found in improved student attendance and report card grades in English, but not in promotion rates or standardized test scores. The effects, although sizable, are not sufficient to neutralize the academic risks with which students enter the program. The positive results are primarily attributed to three of the seven sites. Some evidence supports interpretations that although strong one-on-one mentoring is not an essential component of an effective program that uses outside adults to assist at-risk middle school students, the RAISE model is much more likely to show positive results when one-on-one monitoring is strongly implemented. Success may also depend on the size and composition of the student group to be served. Issues are raised about roles and responsibilities of adult advocates or mentors. Three tables of statistical data and 10 references are included. (Author/SLD)