國家華語文寫作能力測驗兩年前正式上路，以測驗的種類來看，寫作測驗屬主觀性測驗的一種，而主觀性測驗的重要關鍵，即在於評分者一致性的信度要求。對大型標準化測驗而言，主觀性測驗評分的實施，必然以評分規準為準，評分者依評分規準對應試者的反應內容進行評分。以現今臺灣國家華語文測驗而盲，主要分為基礎級、進階級與高階級。寫作測驗的分級，主要是依據分級能力指標，表現出語文能力的高低。然而不同等級的寫作測驗，評分者所依循的評分規準卻大致相同，反應出能力指標與評分規準的落差，進而可能造成評分的理論基礎不足與評分偏誤。本文即擬由評分規準與能力指標之關係入手，以基礎級與進階級記敘文類型為具體對象，探討評分規準與能力指標的在理論及實際的聯繫與對應情形。

The Writing Test of Chinese as a Foreign Language was initiated two years ago. The writing test is one of subjective evaluation which demands to consistent credibility of evaluators. As formal standardized test, practice of subjective evaluation has to fellow standard evaluation as criterion that evaluator assess examinee's answer according to evaluation criterion. Taiwan TOCFL has three different levels, beginner, and learner, superior. Evaluation of writing test is basically according to TOCFL competence indices. However evaluator fellow the similar standard on various levels which reflect discrepancy between TOCFL competence index and evaluate standard, causing deficiency of fundamental theory of evaluation. The paper star with a goal of connection between TOCFL competence index and evaluate standard, beginner writing and narration of learner level as the study object. Then the comparative and connection between, both in theoretically and practically, TOCFL competence index and evaluate standard are discussed.