本文以台灣公司治理評鑑制度為研究標的，探討宣告公司治理評鑑資訊是否可以降低資訊不對稱。本文根據過去文獻，以股市的股票買賣價差做為衡量資訊不對稱的代理變數，並控制其他非資訊不對稱的相關變數。實證證據顯示平均而言，股市的股票買賣價差在宣告日及宣告日前一天都會減少1%。本文也進行長窗期的檢測，實證結果發現對於宣告日前四天至宣告後兩天的期間來說，公司治理評鑑資訊會降低資訊不對稱，然而，當本研究使用交易量作為資訊不對稱的代理變數時，並沒有發現前述結果。本研究也進一步測試投資人的預期行為及公司治理評鑑結果的狀態是否會影響資訊不對稱，結果發現不論前一期公司的EPS是正數或負數，抑或是公司治理評鑑狀態是升等、降等或是不變，宣告公司治理評鑑資訊造成資訊不對稱下降情況仍存在。最後，本文實證結果可以作為立法者、公司管理階層及整個股市參考之用。

This research takes Taiwan’s corporate governance evaluation exercises as the subject to investigate whether releasing their rankings decreases information asymmetry. I follow prior studies and focus on the stock market’s bid-ask spread as a measure of information asymmetry, while controlling for its non-information asymmetry-related components. Evidence herein demonstrates that, on average, the bid-ask spread decreases by approximately 1% on the announcement date of corporate governance evaluation rankings as well as one day before the announcement. I also explore long-window effects, documenting results that releasing corporate governance evaluation rankings does decrease information asymmetry over the four days prior to the announcement and two days after the announcement, but I do not find consistent results when I employ the number of transactions as an alternative measure of information asymmetry. I further examine whether investors’predictable behavior and the status of corporate governance evaluation results influence information asymmetry, offering findings that information asymmetry really diminishes no matter whether the firm’s EPS in the previous period is positive or negative and whether the status of CG evaluation is upgrade, downgrade, or no change. Lastly, I highlight the empirical findings’implications for regulatory agencies, corporate management, and the overall capital market.