專利計量已經成為科技政策研究領域的重要研究工具，透過大樣本的專利引用資訊分析，用以衡量及觀測企業、產業甚至於國家的科技發展程度，抑或用以輔助國家科學技術投入之預算配當議題。然而，除了專利間的相互引用，越來越多學者傾向使用專利對非專利文獻（Non-Patent Reference，NPR）的引用（例如，專利對科學文獻的引用）進行科學技術互動之分析。本研究則旨於探討非專利文獻引用之相關指標的應用評估，以提供讀者在使用非專利文獻進行分析前一整體及全面性的理解。首先本研究整理並回顧過去學者對非專利文獻引用之指標探討，並且針對專利引用及非專利引用之行為策略性思維及行為進探討。本研究更進一步針對美國專利資料庫進行非專利文獻引用的統計分析，進行美國專利引用非專利文獻之應用評估，以及研究者應注意哪些應用非專利文獻進行分析時可能存在的謬誤。 Patent Bibliometrics has been becoming an important approach for Sci-Tech policy research in modern knowledge-based society. For example, large-scaled patent citation analysis can be used to measure Sci-Tech development for enterprise, industry or country as well as to rationalize funding allocation for National Sci-Tech development. In addition to patent reference (PR) usually studied in patent citation research, non-patent reference (NPR) has also become of great importance in literatures. Recently, the popular study of scientific journal citation, which is one type of NPR, in patent is a good example of investigating interaction between science and technology. This study aims to assess the application of NPR in order to provide a complete and comprehensive picture for NPR before it can be adopted as an proper indicator. This study first reviews literatures in relation to the use of NPR as indicator, then investigates the strategic thinking and action based on which PR and NPR are used, finally a large-scaled statistics on NPR in patents published by USPTO is conducted to assess the appropriateness of the use of NPR as indicator. It is concluded in this study that mistake can be easily made if NPR-based indicator is not carefully adopted.