高速鐵路因為具有安全、快速且大量運輸旅客之特性，串起了城市與城市間更為緊密的溝通，目前已經成為各國政府在規劃其運輸網路的重要選項。由於興建高速鐵路的成本較一般鐵路高出許多，如何有效率地透過營運後的績效表現來盡速回收其在興建時期的投資成本，考驗著各國高速鐵路公司的營運管理能力。本文首先以國際鐵路聯盟（UIC）所公布之高速鐵路會員國進行資料蒐集，並參考相關文獻找出投入項變數、產出項變數、及環境變數以做為評估基礎。除了使用傳統資料包絡分析法來對世界各國高速鐵路公司之營運績效進行評比外，亦使用了三階段資料包絡分析法來排除環境因素對於受評單位可能帶來的干擾，以求算出更為公允之技術效率值。由分析結果來看，台灣高速鐵路公司三年平均之技術效率與純技術效率值雖然皆位於荷蘭與日本之後，但仍處於領先族群，優於南韓、德國、及法國等國。此外，我們比較排除環境因素前後之差異後發現，效率值確實有顯著之差異，例如排除環境變數導致之影響後，台灣高速鐵路公司三年之效率值皆為100%，顯示本文所選取之環境變數對於高速鐵路公司之營運效率有重大之影響，此項結果可做為其他即將投入高速鐵路產業之國家作為整體環境評估的參考。

Due to the characteristics of safety, high speed and capability of transporting high volume of passengers, high-speed railway has shortened the length of physical geography and played as a role of connectors between cities. It now has become the preference of governments around the world to design its transportation network. However, the cost of constructing a high-speed railway is very expense, how to turn into a benchmarking company is a very critical issue for every manager of high-speed-rail companies. This paper first applies the traditional data envelopment analysis (DEA) model, then emploies the three-stage DEA models to assess the performance of major high-speed-rail companies in the world. The effects of environmental factors to the performance of high-speed-rail companies are examined as well. Results from empirical analyses show that even the three-year average technical efficiency and pure technical efficiency of the Taiwan High Speed Rail Corporation (THSR) fall behind Netherlands and Japan, it is still better than South Korea, Germany, France and other countries. In addition, it is discovered that when environmental factors are excluded from the DEA model, the resulting performance scores are significantly better than the situation when they are included in the model. This indicates the environment variables we suggest in this paper have a significant impact on the operating efficiency of a high-speed-rail company.