資策會在2005年的調查中，發現四千多位受訪的網友中，最常使用的網路付費服務是拍賣，而許多文獻也指出「信任」是拍賣網站的關鍵成功因素之一。 本論文旨在探討影響網站信任的因素,並以Dayal等人(1999)提出的「信任的金字塔」為主要架構，收集相關的變數製作問卷；進行網路問卷後，回收有效問卷506份。回收資料以SPSS統計軟體進行因素分析，將45個變數以最大變異法直交轉軸，歸納出九個因素並探討信任度因素的重要性，經成對樣本T檢定分析後，將九項信任度因素分為三個等級。 針對使用者特質變數分析信任度因素之影響程度，發現「保護政策」、「評價機制」和「身分認證機制」的影響程度在性別上有差異；而所有信任度因素在年齡與教育程度上都沒有顯著差異。 在網路使用者習性變數方面，「消費者間互動」的影響程度在每日上網時間上有差異；「評價機制」的影響程度在交易次數上有差異；「消費者間互動」及「網站服務品質」的影響程度在交易金額上有差異，而除了「身分認證機制」外，其他信任度因素之影響程度皆與滿意度成正相關。 根據使用者最常使用的國內拍賣網站來進行分類，發現Yahoo！奇摩拍賣的使用者特質與本研究整體樣本的特質一致。另外，針對問卷中影響信任度較高的機能項目，實際調查國內拍賣網站的執行情形，發現執行比率有偏低的現象。 In 2005, a more than four-thousand sample survey held in the MIC of III showed that the most popular payment internet service was online auction. Besides, many literatures also proved that “trust” was one of the critical success factors of online auction. The primary objective of this study is to investigate factors influencing consumer trust on the website. Based on the framework of「The Trust Pyramid」(Dayal、Landesberg and Zeisser, 1999), a questionnaire was developed by gathering related variables from previous literature review. Put this questionnaire on the BBS, invited Internet users to fill in it and obtained 506 valid samples. Using factor analysis program of SPSS by applying Varian rotation principle component analysis, 45 variables are reduced to 9 factors. In the same time, we used Paired Samples T-Test to classify these trust factors into three grade levels. Some statistical tests were executed between user demographical variables and trust factors. We found that the gender can influence the trust factors of “Protective policy”、“Feedback” and “Authentication”. However, there were no significant differences between the trust factors and the age as well as education of users. In the variables of users’ behavior on the Internet, the influence degree of “Consumer collaboration” had difference in daily surf time; the influence degree of “Feedback” had difference in number of transactions; the influence degree of “Consumer collaboration” and “Fulfillment” had difference in transaction value. In the influence degree of the trust factors had positive correlation with users’ satisfaction except the factor of “Authentication”. The most popular local online auction websites of this survey samples were classified and found that the characteristics of users of Yahoo were consistent with the whole sample of this survey. Besides, we identified the websites mechanisms that could influence users’ trust factor and found that the implementation ratios of these mechanisms were still low on local auction websites.