分級字表是華語文教學標準體系的一環，與能力指標、分級詞表及分級語法點表等形成一評估閱讀及語言學習成效的重要體系。目前臺灣對外華語教學參考資源僅有華測會提供的8,000詞詞表，並無統一的字表供教學、編輯教材或測驗評量等使用；國外字表又各有地域性，不適用於臺灣，因此，本文針對華語學習者字表，探討其分級、建置的方法及成果。首先，釐清字、詞表分別建置的原因、教學字表與一般字表的差異，及母語學童字表與華語學習者字表的差異等；其次，說明建置華語學習者字表需考量的因素，依流程逐步介紹建置字表所運用的統計法和專家法。本文5位專家調校信度在Weighted Kappa計算中達「幾乎完全吻合」等級，調校幅度約為3成，顯示字表以語料庫統計法為主；最後，本文提出字表研發成果，以常用度、語義類聚及書面語╱口語相對字頻比等3種方式呈現，並以語料覆蓋率與國外字表比較分析，驗證本字表為最能代表臺灣當代用字現況的字表，在基礎級的語料覆蓋率上，平均高出其他字表4個百分點以上。整體而言，本文建置的字表具有客觀、可再現、在地性及符合二語學習者需求等特色。

Graded Chinese character lists, word lists, grammatical points lists and language proficiency indicators form a standards system, which is usually adopted to evaluate the effectiveness of reading and learning of Mandarin Chinese. As there is not a localized Chinese character list available and Chinese character lists used overseas cannot represent the actual usage of Taiwanese, in this paper, we propose a methodology of constructing a graded Chinese character list for L2 learners. First, we discuss methodological issues concerning word lists and character lists and their applications in different contexts. Second, we point out the factors that influence the construction of character lists, including statistical factors and expert-based factors, and then describe our building database procedures in details. Finally, for convenience of application, the result of the graded character list is arranged according to three types of categories: frequency, semantics, and the written/spoken corpus character ratio. The study concludes that the reason for our character list to be able to achieve the best performance of corpus coverage is because the list is objective, reproducible, localized, and as such it meets the needs of second language learners.