Background: Alexithymia, an inability to identify or describe emotions, is associated with suicidality yet the correlation with single or repeated suicide attempts is less clear. Aims: We aimed to assess the modifiability of alexithymia following a group psychosocial intervention focused on improving emotional literacy in those with a history of recurrent suicide attempts (RSA). Method: A total of 169 participants with self-reported RSA completed pre- and postgroup assessments of a 20-week group therapy intervention. Questionnaires assessed alexithymia, depression, impulsivity, and hopelessness; the Toronto Alexithymia Scale (TAS-20) was the primary outcome. Data were analyzed using multiple imputation. Results: Participants had on average 7.8 lifetime suicide attempts, 73% were female, and 16.6% had a >13-point reduction in TAS-20 scores after 20 weeks. Directed acyclic graph (DAG) analysis demonstrated significant relationships between alexithymia, depression, hopelessness, problem-solving, and satisfaction with life. Age of onset of suicidality was the only factor predictive of postintervention TAS-20 score in univariate linear regression. Limitations: The study limitations were its sample size, insufficient resources, and missing data. Conclusion: A change in TAS scores indicated that alexithymia can be a modifiable treatment target. Being able to identify and describe feelings may lead to improvement in depression, hopelessness, problem-solving, and satisfaction with life in this population.