Objective: In a recent trial, moral reconation therapy (MRT)-a cognitive-behavioral intervention for criminal recidivism-was not more effective than usual care (UC) for veterans in behavioral health treatment. To determine for whom treatments of recidivism are most effective, we tested if recency of criminal history or psychopathic traits moderated MRT's effects on outcomes. Method: In a multisite trial, 341 veterans (95.3% male; 57.8% White/Non-Hispanic) with a criminal history who were admitted to behavioral health treatment programs were randomly assigned to UC or UC + MRT and followed at 6- and 12-months. Incarceration (yes/no) or criminal conviction (yes/no) in the year prior to enrollment and psychopathic traits at baseline (median split) were prespecified as moderators of treatment effects on primary (criminal thinking, criminal associations) and secondary outcomes (legal, employment, and family/social problems; substance use problems and days of use). Results: Among veterans incarcerated in the year prior to enrollment, MRT (vs. UC) was associated with greater reductions in criminal associations (6 months) and days drinking or using drugs (12 months). Among those convicted in the year prior to enrollment, MRT (vs. UC) was associated with greater reductions in employment problems (12 months) and days drinking or using drugs at each follow-up. For those high in psychopathic traits, MRT (vs. UC) was associated with greater reductions in days drinking or using drugs at each follow-up. Conclusions: For veterans in behavioral health treatment with recent criminal histories and high in psychopathic traits, MRT may be effective for reducing risk for criminal recidivism. What is the public health significance of this article? This study found that for veterans in behavioral health treatment programs with a criminal history, a group treatment aimed at reducing risk for criminal recidivism is effective for those with more recent criminal histories and those with greater psychopathic tendencies.