Objective: To explore and test a new blending prospective payment that works for integrated care delivery system. Methods: Referring to Accountable Care Organization and domestic reforms, we designed a performance-based and prospective global budget payment mechanism which mainly contained strategies as medical alliance contract, prepay by DRGs and performance-based management. Through a quasi-experiment (2 towns in control and treated group respectively, and lasted from 2012-2014), we tested its effects on controlling the inpatient spending and continuity of care. Results: There were 38,980 cases included from inpatient claims data and 194 medical records that from township- and county hospitals. Compared to control group, the average hospitalization rate in treated group significantly declined by 0.08%, the likelihood of using upper level hospitalization significantly declined by 0.2%, and the continuity of care significantly increased by 33.8%. However, the effects of new model might be underestimated by the imperfect implementation of compulsory referral system. Conclusion: The medical alliances should center on combined objectives as stimulate medical cooperation and improve on quality of care. They also should make the blending prospective payment decisions on basis of information-shared grading and referral medical system and empirical evidence