OBJECTIVES: To improve the outcome of stroke care in rural areas, a model of Stroke Unit Equivalent Care (SUEC) with/without early supported discharge (ESD) services was developed in Alberta, Canada. The objective of this study was to evaluate the cost-effectiveness of SUEC and SUEC followed by ESD (SUEC&ESD) compared with the pre-setting (nonSUEC). METHODS: We developed a three-arm cost-effectiveness model to compare three stroke care models: nonSUEC, SUEC and SUEC&ESD using health administrative data. We estimated the 180-day mortality in a generalized linear model with binomial family and estimated costs with gamma family. We applied the predicted mortality to Alberta age-sex specific life expectancy to calculate life-year expectancy for each patient after stroke onset. We applied a multi propensity score method with multinomial logit model to correct potential bias from observation data. We calculated the risk-adjusted incremental cost (IC), incremental effectiveness (IE), and then incremental cost-effectiveness ratios (ICERs). We performed 2,000 bootstrapping on study populations and presented them in cost-effectiveness plane for sensitiveness analysis. RESULTS: Of 988 patients retrieved, 895 new stroke patients were selected for the analyses. SUEC was less costly and more effective than the nonSUEC (IC= -$3,378, IE=0.45 year, ICER= -$7,522 per life-year gain). The SUEC&ESD was more costly but also more effective than either the SUEC (IC=$5,179, IE=1.64 year, ICER=$3,154 per life-year gain) or the nonSUEC (IC=$8,557, IE=1.19 year, ICER=$7,174 per life-year gain). CONCLUSIONS:  With a willingness-to-pay value threshold of $50,000, the SUEC was dominant strategy with lower in-hospital mortality and lower direct health care costs than the nonSUEC. The SUEC&ESD was also cost-effective when compared with the nonSUEC or the SUEC. This new model of care reduces urban/rural care discrepancies providing lower mortality for stroke while providing good value to the healthcare system.