This 5-year longitudinal randomized study focused on the effects of a dropout prevention program implemented at two urban high schools. The program incorporated the characteristics of personalization found in previous research to have some positive effects, and typical of the type of programs implemented in many high schools nationwide. The primary program component was the provision of a facilitator for about 60 program students at each school to encourage students regarding attendance, their academic coursework, and personal issues. In particular, the study examined whether at-risk students randomly assigned to the program had better outcomes (particularly attendance, on-time promotion in grade, and high school completion) than those at the same high schools that were assigned to the control group. The article explores reasons for the nonsignificant effects of this program, emphasizing the need for earlier intervention prior to the ninth grade year among at-risk urban students. (Contains 9 footnotes and 9 tables.)