Rates of alcohol and/or substance use (ASU) among residents of predominantly Black and marginalized communities are similar to ASU rates in White communities. Yet ASU has worse consequences in predominantly Black and marginalized communities (e.g., higher incarceration). We randomized participants to one of 16 intervention conditions using a 2 full factorial design to optimize a multilevel intervention reducing ASU among 602 formerly incarcerated men with substance-use-disorders (SUD). Candidate intervention components included (1) critical dialogue (CD; six weekly 2-hour-long group sessions vs. no CD sessions), (2) Quality of Life Wheel (QLW; six weekly 1-hour-long group sessions vs. no QLW sessions), (3) capacity building projects (CBP; six weekly 1-hour-long group sessions vs. no CBP sessions), and (4) delivery by a trained peer versus licensed facilitators. Outcome was percentage of days in which participants used alcohol, cocaine, opioid, and/or cannabis in previous 30 days. Intent-to-treat analysis did not meet a priori component selection criteria due to low intervention attendance. After controlling for intervention group attendance (percentage of sessions attended), peer-delivered CD and CBP produced statistically and clinically significant main and interaction effects in ASU over 5 months. Per the multiphase optimization strategy framework, we selected peer-delivered CD and CBP for inclusion as the optimized version of the intervention with a cost of US$1,380 per 10 individuals. No adverse intervention effects occurred. CD and CBP were identified as the only potentially effective intervention components. Future research will examine strategies to improve attendance and test the optimized intervention against standard of care in a randomized-controlled-trial.