According to the costly arbitrage model of Pontiff (2006), idiosyncratic risk increases holding costs that limit arbitrage. Since short sales are actively used for arbitrage, the costly arbitrage model should apply to short sales, which is the case according to the empirical findings of two recent studies: Au, Doukas, and Onayev (2009) on US markets, and Duan, Hu, and McLean (2010) on U. K. markets. Both studies offer evidence that idiosyncratic risk acts as a deterrent to short sales and affects returns subsequent to short sales. This study investigates whether a similar pattern exists in the Korean equity market using short sales data for stocks listed on the Korea Exchange from July 2004 to June 2011. Our study contributes to the literature in three important ways. First, it is the only study ever conducted outside the U.S. and U.K. markets on the relation between idiosyncratic risk and short sales. Second, different from Au et al. (2009) and Duan et al. (2010), we use short sale volume data rather than short interest data in our analysis. Using short sale trading volume is preferable to short interest volume because the former captures short sales activity more accurately than the latter, which often under- or over-estimates short sales activity. Third, we perform an investor-type analysis. Our data enable us to distinguish short sale volume by different investor types: individuals, domestic institutions, and foreigners. An investor-type analysis offers important insights into the effects of idiosyncratic risk on short sales activity. This is because different types of investors have different motives for short sales, an issue closely related to whether a certain type of investor engaging in short sales is more informed than other types. How idiosyncratic risk affects short sales made by different types of investors, and how it is related to the information content of short sales made by certain types of investors, are interesting questions. We measure short sales activity by short volume ratio (SVR), defined as the cumulative daily short volume of a stock in a month divided by the total trading volume of the stock in the same month. Following Ang, Hodrick, Xing, and Zhang (2006), we estimate the idiosyncratic risk of a stock in a given month from the Fama-French three-factor model. Our analysis is performed in two ways: a comparative analysis based on 20 groups formed by levels of idiosyncratic risk and SVR (5 by 4), and Fama-MacBeth regressions of (1) SVR on idiosyncratic volatility and (2) abnormal returns on SVR and idiosyncratic volatility. Both regressions use control variables that are believed to affect short sale volume or returns. Our findings can be summarized as follows. First, we find a negative relation between idiosyncratic risk and SVR, which is consistent with the recognition by investors of idiosyncratic risk as an arbitrage cost that limits short sales activity. Second, short sales made by domestic institutions show the strongest negative effect of idiosyncratic risk on SVR, followed by short sales made by foreigners. We do not find evidence of such a negative relation between idiosyncratic risk and short sales made by individuals. These patterns are consistent with different motives for short sales made by different types of investors. Institutions use short sales primarily for hedging and arbitrage. Hence, idiosyncratic risk is a significant binding factor. It binds short sales made by individuals less because they use short sales mainly for speculative purposes. Third, the greater the SVR, the lower the returns after short sales. This suggests that short sales predict future stock returns and that, on average, short sale traders are well informed. Fourth, the negative relation between SVR and future returns is stronger for firms with greater idiosyncratic risk. This finding is consistent with the model of Diamond and Verrecchia (1987), which states that costly short sales are used by informed traders. Finally, the effect of idiosyncratic risk on the relation between SVR and future returns is stronger for short sales made by foreign investors–a finding that suggests an informational advantage of foreigners over domestic institutions and individuals.