The purpose of this study is to analyze the impact of CEO turnover announcements on abnormal stock returns and the determinants of abnormal returns, using a sample of 159 Korean listed firms that disclosed CEO changes between January 1, 1993 and December 31, 1997. The empirical findings can be summarized as follows. For the subsample where the predecessor was not appointed as CEO of another company, stock prices declined significantly by -0.880% during the announcement period (AD-1 to AD+1) and by -1.746% during the period from announcement day to eight days after (AD-1 to AD+8). These stock price declines were statistically significant at the 1% level. In contrast, for the subsample where the predecessor was appointed as CEO of another company, stock prices exhibited no notable pattern during the announcement day or various periods surrounding it. These results suggest that investors perceive turnovers where the predecessor is not appointed as CEO elsewhere as disciplinary turnovers, and they evaluate such disciplinary turnover announcements as negative information indicating that managerial performance at the time of announcement is worse than expected, rather than as positive information that a poorly performing manager has been replaced and stock prices will rise in the future. Conversely, investors perceive turnovers where the predecessor is appointed as CEO of another company as reward-based turnovers, and they do not evaluate such reward-based turnover announcements as negative information suggesting that managerial performance is worse than expected. Meanwhile, cross-sectional regression analysis conducted on the full sample to identify determinants of abnormal stock returns on the announcement day yielded the following results: the predecessor CEO's managerial performance, a dummy variable representing the predecessor's status, and firm size were statistically significant. However, the successor's ability and background were found to be unrelated to abnormal returns.