OBJECTIVE:This study examined long-term outcomes (at 36 months) from Washington State's policy of expediting Medicaid enrollment for prison releasees with severe mental illness and compares them with previously reported short-term outcomes (at 12 months). METHODS:Linked administrative data on prison releasees (2006-2007) were analyzed by using a quasi-experimental design comparing those referred to expedited Medicaid (N=895) with a control group of those not referred (N=2,189). Aggregate outcomes were analyzed with inverse probability of treatment-weighted logit models. RESULTS:Expedited Medicaid had a sustained effect on both increased months of enrollment (p<.01) and increased use of community mental health and general medical services (p<.01) 36 months after prison release. However, expedited Medicaid did not reduce criminal recidivism, consistent with 12-month findings, Conclusions: Outcome results at 12 months were sustained at 36 months-namely, expedited Medicaid for released prisoners with severe mental illness improved enrollment and service use with no effects on criminal recidivism.