Abstract The current two-study program of research sought to 1) quantitatively summarize existing empirical knowledge of the potential relationship between mental health and recidivism among justice-involved women using meta-analysis, and 2) make a novel contribution to the development of mental health focused correctional profiles for a population of custodial women in Maine. Study 1 was a meta-analysis including all studies of justice-involved women examining the relationship between any aspect of mental health and recidivism. Eighteen studies met inclusion criteria following a comprehensive literature review. The quantitative results suggested that depression, PTSD, psychiatric history, and the presence of any mental disorder relative to no mental disorder were all independently associated with significant and modest to moderate increases in recidivism rates. Conversely, anxiety, psychotic, and unspecified personality disorders, as well as self-harm/suicidality were not significantly related to recidivism. Study 2 used latent class analyses (LCA) to identify and explore latent mental health profiles in a sample of 920 women incarcerated in a Maine state prison. This study aimed to determine whether women could effectively be grouped into distinct mental health classes differing in terms of the presence, nature, and severity of mental health challenges. Indicators of mental health were extracted from the Service Planning Instrument for Women (SPIn-W; Orbis Partners, 2006). Results of Study 2 revealed three distinct latent classes that varied in terms of their respective mental health concerns: Class 1 "severe mental health needs" (30%), Class 2 "moderate mental health needs with high rates of self-harm and suicide"(20%), and Class 3 "low mental health needs" (50%). Classes 1 and 2 evidenced signific antly higher recidivism rates relative to Class 3. The main practical utility of identifying such correctional profiles is the continued development of more effective client management strategies, improving service provision and resource allocation, further enhancing the predictive validity of risk assessment, and accurately identifying treatment needs for individual women. Finally, such profiles can be used to link individual clients with the most beneficial treatments and delivery modalities.