This study examines the impact of mental health services on arrestsof offenders with a serious mental Illness (SMI) by assessing changes inassociations between receipt of outpatient and emergency room/inpatientservices and arrests one, two, and three quarters later. A variety ofdata sets were used for identifying 3,769 offenders who were in thePinellas County Florida jail between 7/1/2003 and 6/30/ 2004, and 7,755offenders who were in the Harris County Texas jail between 10/1/2005 and9/30/2006. Arrests, out-patient and emergency room/inpatient serviceswere assigned to one of 16 ninety-day periods between 7/1/2002 and6/10/2006 in Pinellas County and one of 12 such periods between10/1/2004 and 9/15/2007 in Harris County. Generalized estimatingequations were used. Covariates were age, gender, race, diagnosis, andhomelessness. The results were also adjusted for exposure to arrests. InPinellas County, outpatient services significantly reduced the risks ofarrests 1 quarter later by 17% (odds ratio [OR] = 0.83, 95% confidenceinterval [CI]: 0.78-0.87, p < .001), two quarters later by 11% (OR =0.89, 95% CI: 0.84-0.94, p < .001), and three quarters later by 9%(OR = 0.91, 95% CI: 0.86-0.96, p = .001). In Harris County, theseservices reduced the risk of arrest 1 quarter later by 5% (OR = 0.95,95% CI: 0.91-0.99, p = .028), but not two and three quarters later. InPinellas County, ER/inpatient services increased the risk of arrests by22% (OR = 1.23, 95% CI: 1.15-1.30, p < .001), 8% (OR = 1.08, 95% CI:1.02-1.15, p = .010) and 11% (OR = 1.11, 95% CI: 1.02-1.16, p = .001)one, two, and three quarters later. In Harris County, these servicesincreased the risk of arrest only 1 quarter later (OR = 1.16, 95% CI:1.11-1.22, p < .001). Results suggest that service receipt and itstiming may have had some impact on the arrests of adults with a SMI andcriminal justice involvement. (Authors).