Objective:This analysis investigated whether persons with serious men-tal illnesses have longer jail detentions than other detainees andwhether they are released by different legal mechanisms. Methods:Jailrecords and mental health service records from a Medicaid databasewere matched for all admissions to the Philadelphia jail system in 2003.Survival analysis techniques were used to compare length of jail stays ofpersons with and without serious mental illnesses (N=24,290). Seriousmental illness was defined as a diagnosis either in the schizophreniaspectrum (DSM-IVcode 295.XX) or of a major affective disorder (DSM-IVcode 296.XX) recorded in Medicaid records (2001–2003). Mecha-nisms of release were also examined for those with release dates beforeSeptember 1, 2005 (N=20,573) Results:Just over 50% of the 1,457 per-sons with serious mental illnesses were released from jail within 30 daysof incarceration, compared with 56% of the other detainees. Mental ill-ness status was not found to be a significant predictor of longer deten-tions. Forty-nine percent of those with serious mental illnesses were re-leased from jails through unpredictable release mechanisms, such asbail, release from court, or withdrawal of a bench warrant, whereas only19% were released through mechanisms that had release dates that al-lowed adequate time for discharge planning. Conclusions:The findingssuggest that reentry programs and other jail-based interventions forpersons with mental illnesses should ensure that they have the capacityto rapidly identify and serve clients with shorter and more unpre-dictable stays or risk not being responsive to the needs of a substantialproportion of this population. (Psychiatric Services61:458–462, 2010)