Objectives: Little research has assessed the nature of veterans' departures from permanent supportive housing (PSH), which may be of positive valence (e.g., moving into more independent housing). This study aimed to identify participants appropriate for "graduation" from PSH and how to support their transitions. Method: This mixed methods study used qualitative data from PSH program staff, 445 PSH participants' responses to a survey assessing their experiences and administrative records, and qualitative data from a subsample of 10 participants who graduated from the program. Participants were classified as "stayers" (retained in PSH for at least 600 days); "graduates" (exited with improvement in community integration); or "involuntary leavers" (exited for reasons of negative valence). Template analysis of qualitative data from PSH staff described graduation processes; qualitative data from participants were analyzed using a thematic analysis approach. The study compared veterans' characteristics using chi-square and analysis of variance (ANOVA) tests; a multinomial logistic regression assessed correlates of graduates' and involuntary leavers' exits from PSH. Results: Approximately one half of participants who exited the program were graduates. Processes used by program staff to identify potential graduates varied. Participants' self-report of substance use and mental health problems was associated with involuntary leaver status. Frequency of case management, a trusting relationship between participant and case manager, and participants' receipt of compensation related to disability incurred during military service were associated with graduation. Conclusions and Implications for Practice: To support successful transitions from PSH, programs should focus on providing high-quality case management that may respond flexibly to participants' varying recovery needs.