This randomized, controlled study examined the effectiveness of two mentoring programs, child mentor relationship training (CMRT) and peer assistance and leadership (PAL®), on high school mentor empathic behaviors and child mentee behavior problems. Participants were 60 young, at-risk students (61.7% male; 38.3% Hispanic/Latino/a, 31.7% Caucasian, 21.7% African American, 8.3% biracial) and 30 high school students (53.3% male; 66.7% Caucasian, 26.7% Hispanic/Latino/a, 0.03% African American, 0.03% Asian). Mentors and mentees were randomly assigned to CMRT or PAL®, which was treatment as usual in the participating school district. Results from 2 (group) by 2 (time) repeated measures ANOVAs indicated compared to the PAL® treatment group over time, mentors in the CMRT group demonstrated statistically significant improvement in empathic behaviors with a large treatment effect, as rated by independent observers. Analysis revealed a moderate treatment effect with CMRT group mentee behavior problems, but the difference was not statistically significant between treatment groups over time. Further analysis revealed the CMRT group demonstrated statistically significant reductions in behavior problems from pre- to post-test with a very large treatment effect. Overall, findings support CMRT as a promising school-based intervention for at-risk young children that potentially increases school counselor efficiency.