In both counties involved in the study, having Medicaid benefits available at release was linked to a 16-percent reduction in the average number of subsequent detentions of severely mentally ill detainees. After the analysis controlled for demographic and clinical variables, more days on Medicaid were linked with a reduced number of subsequent detentions in King County (Washington State) and more days in the community before subsequent arrests in both counties (King County and Pinellas County, FL). No association was found between Medicaid status and the severity of the offense leading to rearrest in either county. Further research is required in order to determine ways in which greater reductions in reoffending can be achieved for jail detainees with severe mental illness. The study used a quasi-experimental design in which administrative data were collected and analyzed. Subjects of the study were all persons with severe mental illness released over a 2-year period from jails in King County (n=5,189) and Pinellas County (n=2,419). They were monitored for 12 months after their release from jail. Those who were receiving Medicaid benefits at release and those who did not receive these benefits were compared on three variables: number of arrests after release, how soon any rearrests occurred, and the seriousness of offenses linked with any rearrests. The data were analyzed with negative binomial, Cox proportional hazards models, and logistic regression with adjustment for dependent observations. 3 tables and 32 references