The return of hedge funds often has different statistical characteristics and risk return relations from those of regular investments due to the fact it adopts some flexible investment skills in practice． Therefore，it is difficult to evaluate the performance of hedge funds by using some traditional methods． Based on LASSO quantile regression，the paper identifies some important risk factors，which are helpful to explain hedge fund investment style，and provides a new method by considering the relationship between return and risk across quantiles to evaluate the performance of hedge funds． For illustration，a style portfolio of hedge funds is constructed through LASSO quantile regression． Then，the style portfolio is compared with some classical methods for portfolio choice，such as mean regression portfolio，equal-weighted portfolio，and Markowitz portfolio． Empirical results show that our evaluation method based on LASSO quantile regression is relatively effective and its portfolio investment scheme may obtain higher risk adjusted returns．