This study employs the quantile regression model to examine the non-monotonic impact of CEO stock-based compensation on firm performance, using the data for U.S. non-financial firms from 1993 to 2005. Furthermore, the quantile-varying
pattern of the impact of stock-based compensation on firm performance is robust after controlling for the industrial and yearly effects. It is also robust to the use of the payfor-performance sensitivity as an alternative explanatory variable or the market-based measure of performance as the dependent variable, or the consideration of the suspected endogenous problem between firm performance and stock-based compensation.