The present study examined possible predictors of youth client retention in therapy in a large community-based sample. We used several conceptualizations of retention, including (a) "intake retention" (i.e., returned to treatment after intake session); (b) "mutual termination" (i.e., termination agreed upon by family and therapist), (c) "mean treatment duration" (i.e., completing the mean number of sessions in the agency), and (d) "total treatment duration" (i.e., total number of sessions). Archival data from over 400 children and adolescents who sought treatment at a large public mental health clinic were analyzed using regression analyses. Although different predictors were identified across the various conceptualizations, a few robust predictors emerged including ethnicity and client symptom severity. Clinical implications and future research directions are discussed.