醫院在高度競爭性的知識基礎經濟市場中，唯有有效整合各種專業知識、積極進行組織創新，才能有效增強其競爭優勢。近來策略聯盟已成為許多醫院經常利用的策略性工具，隨著策略聯盟日趨盛行，醫院的聯盟網絡亦趨多元，因此本研究試圖針對醫院具有的聯盟網絡多元性，對其組織創新的影響進行深入的探討。透過知識基礎觀點，本研究共發展出七大項研究假設。在實證的部份，以郵寄問卷的方式，針對全國460家地區醫院級以上醫院之高階主管或聯盟管理者進行問卷調查，有效問卷回收率為35.45%，在檢驗各樣本代表性與問卷信效度後，本研究進一步以階層迴歸進行研究假設的驗證。分析結果顯示，聯盟網絡多元性對醫院組織創新確實有顯著的倒U型關係，除直接性影響外，其亦會透過外部知識獲取的部份中介來影響組織創新。此外，潛在性吸收能力與IOR知識庫在聯盟網絡多元性對外部知識獲取的影響路徑上，具有完全中介的效果。至於實現性吸收能力方面，其結果發現在外部知識獲取對組織創新的影響路徑上，亦扮演中介性的角色。而本研究另發現醫院自聯盟網絡進行的外部知識獲取，對於醫院管理創新的強化較具有影響力，在提昇技術創新程度上的影響性則較有限。 Hospitals have existed in highly competitive market based on knowledge economy. Under this business environment, hospitals should integrate several kinds of specialties efficiently and drive organizational innovation aggressively to enhance competitive advantage. More recently, strategic alliances have been frequently employed by hospitals as strategic tools. With the increasing usage of strategic alliances, alliance networks of hospitals have become more diverse. Therefore, we attempted to address the impact of alliance network diversity on organizational innovation. Since we argued the influence of alliance diversity would originate from knowledge benefits and limits, we hence adopted knowledge-based view (KBV) to analyze it. Four research constructs generated by the knowledge-based view included inter-organization relationship knowledge store (IOR knowledge store), potential absorptive capability (PACAP), realized absorptive capability (RACAP), and external knowledge acquisition. Based on these constructs, we built up seven research hypotheses. The selfadministered mail survey was employed to collect data. Questionnaires were mailed to 460 hospitals accredited as district hospitals and above. The informants were the top managers or alliance managers of these hospitals. The overall valid response rate was 35.45%. We finally obtained 138 samples excluding hospitals not joining any alliance. After testing sample representativeness, we adopted the confirmatory factor analysis to test the construct validity of each research construct and hierarchical multiple regression to test our hypotheses. Our findings indicated alliance network diversity would affect organizational innovation directly, and the relationship between these two constructs exhibited an inverse-U curve. External knowledge acquisition would partially mediate this relationship in addition to direct effect. Further, PACAP and IOR knowledge store would fully mediate the relationship between alliance network diversity and external knowledge acquisition. RACAP would also play the mediating role in explaining the influence of external knowledge acquisition on organizational innovation. There was one other thing that is important for us. We found that the impact of knowledge gathered by hospitals from alliance networks on management innovation is greater than technological innovation. According to the above research findings, we testified that alliance network diversity would direct and indirect affect organizational innovation of the hospital. Besides, KBV actually was a vital theoretical logic while managers and researchers discussed how to enhance organizational innovation by alliance network diversity.