本文探討研發支出資本化與攤銷之理論，並以國內股票上市公司為實證對象。本研究使用聯立方程式及Almon lag procedure來檢測企業研發投資效益遞延之情形，然後以Feltham and Ohlson（1995）之評價模式，探討研發支出資本化與攤銷後之權益帳面價值與盈餘之特性，並探討研發投資是否為股票報酬之長期風險因素。本研究並以研發強度（研發費用對銷貨凈額比）區分高、低研發強度二組樣本，比較其與研發有關問題。 本文發現如下： (1)研究發展支出效益遞延實現之現象，似乎存續自第三年開始至未來數期（第四、五年）；平均而言，一元之研發投資可於未來4-5年產生2元以上之獲利，高研發強度廠商則有3元以上之獲利；而低研發強度之公司，研發支出獲益則不顯著的於研發初期二年實現。 (2)在股票評價模式中，高研發強度廠商之研發支出採取資本化，對模式解釋能力較高；而低研發比例（強度）廠商之研發支出則採取費用化，「似乎」對模式解釋能力較高。 (3)目前「一般公認會計原則」對於評價模式的解釋，對低研發強度的廠商之解釋能力較高；反之，對高研發強度廠商較不適用。 (4)研發資本存量對市場比率為股票報酬之長期風險因素。 / This study analyzes the theory of capitalization and amortization of R&D expenditures. This study uses the simultaneous equations and Almon lag procedure to examine whether earnings reflect benefits from past R&D expenditures of public firms in Taiwan. Based on asset valuation model generated by Fetham and Ohlson(1995) and Bernard (1995), this study examines the properties of coefficients of parameters of valuation model and explainary power. In addition, this study estimates the R&D capitalization, and tests whether the R&D capitalization is the long risk factor of stock return or not. Moreover, this study groups sample firms into high or low intensity groups by R&D intensity (R&D expenses-to-sales ratio), and compares the above issues of capitalization of R&D expenditures. The major findings of this study follows: (1)Earnings almost reflect realized benefits from R&D, but there are two years time lag. On average, every one N.T. dollar invested in R&D will produce 2 N.T. dollars profits during four or five - years period. In the high R&D intensity group, every one N.T. dollar invested in R&D will produce 3 N.T. dollars profits in future. On the contrary, in the low R&D intensity group, the benefits of R&D outlays are insignificantly reflected in the first two years. (2)The explainary power (Radj2) of valuation model of R&D capitalized in the high R&D intensity group is higher than that of in the low R&D intensity groups. (3)The relevance of stock valuation generated by present GAAPs for the high R&D intensity group is lower than that of the lower R&D intensity group. (4)The R&D capital-to-market values ratio is the long risk factor of stock returns.