An assessment of mentoring outcomes suggests that mentoring may provide some protection against depression among high-risk youth; however, it was less likely to reduce depression among youth already depressed at enrollment. No link was found between mentoring and improvements in either family support or other adult support. There was no difference in the likelihood of being arrested between those youth in the mentoring program and those not in a mentoring program. Youth who had been enrolled in mentoring for at least 6 months were 75 percent less likely to report using marijuana at followup. A profile of the youth who enrolled in the NFBI indicates that they ranged in age from 12 to 19, consisted of an approximately equal number of boys and girls, and had at least one arrest. The crimes they committed ranged from serious crimes against persons to status offenses such as truancy. Many of the youth lived in difficult home environments, with almost one-third residing in public housing projects. A description of the NFBI mentoring programs addresses one-to-one mentoring, group mentoring, and team mentoring. Although the mentoring program models across sites differed, all sites made matches on the basis of gender and common interests. Whenever possible, the sites took into account the skills and experiences the mentors would bring to the relationship with a particular youth. All sites asked mentors to commit to involvement with the youth for a minimum of 1 year and to meet with the youth a minimum of 1 or 2 hours a week for 4 hours a month. All of the sites provide regular support to their mentors. 7 tables, 1 figure, 37 references and appended response rates and survey instruments