The purpose of this research was to investigate the additive therapeutic effects of a yoga intervention on the anxiety, depression and behavioral problems of conduct-disordered male adolescents in residential treatment. In addition, the moderating effects of callous-unemotional (CU) traits on outcome measures were assessed. The program consisted of a four-week intervention program in which participants were randomly assigned to either the yoga group (n=25), in which they practiced yoga with an instructor, or the control group (n=19), in which they met for a supervised study hall. The study included pre-testing on symptoms of anxiety, depression and CU traits, and post-testing on anxiety and depression measures only. Behavioral data were unavailable due to unanticipated program changes. A repeated measures MANOVA was utilized to investigate the benefits of yoga practice on a combined mental health variable that consisted of two dependent variables, anxiety and depression. A significant effect for time, but not for the interaction between time and group, was found. This indicated that both groups' scores decreased over time on the depression and anxiety variables, but that there was no statistically significant difference between the treatment groups' depression and anxiety scores over time. In spite of non-significant results, additional exploratory analysis was conducted. Results indicated a trend towards significantly greater decreases in anxiety outcomes for the yoga group vs. the control group over time. The moderating effects of CU traits on the relationships among the treatment conditions and anxiety outcomes were found to be non-significant. Limitations of the present research, including low sample size and statistical power, are discussed.