In the past, research on evaluating the portfolio efficiency of mutual funds was mostly based on the mean-variance framework proposed by Markowitz. However, for Chinese short-term mutual funds (CSTMF), short-term mutual funds with different lock-up periods will automatically roll over after the lock up period to restart the funds unless instructed to terminate. By reducing the liquidity of their investments, the lock-up period generates an opportunity cost for investors. The liquidity impacts of different CSTMFs will depend on their lock-up periods. In this paper, we evaluate the portfolio efficiency of the funds by justifying using the lock-up period as a measurement to measure the liquidity risk faced by the investors from the perspective of investors. Therefore, here, we propose to use the return-variance-liquidity framework to evaluate the performance of the funds. Furthermore, we select 28 CSTMF funds to conduct empirical analysis with appropriate DEA models. Firstly, we carry out some comparisons with the traditional mean-variance framework. The results show that the length of the lock-up periods has a significant impact on the fund’s performance, which has important guidance for the investment strategy. Finally, we conduct the benchmarking analysis for the 28 funds and introduce a new DEA model, present how to reduce liquidity risk under the premise of given return and variance, and provide valuable investment advices by combining these with sensitivity analysis. Overall, there are three conclusions: (1) most inefficient funds can be improved using portfolios consisting of funds with shorter lock-up period. (2) For some inefficient funds with shorter lock-up period, the current return and variance levels can be achieved by a portfolio. (3) Among the funds with higher efficiency scores, funds with shorter lock-up period have higher stability.