Codend size selectivity is the keystone for management of trawling fisheries and sustainable exploitation of fisheries resources. In order to assess the size selectivity of trawl codends, experimental sea trials are often conducted and fishing data is manipulated using some mathematical models to obtain selectivity parameters, such as 50% retention length (L50) and selection range (SR), and selectivity curves. Providing the values of this information, however, is highly inadequate. It needs to account for the uncertainties in size selectivity and represent them in confidence intervals. To do this, the double-bootstrapping technique is intensively applied in fishing gear selectivity studies. In this study, the basic principle of the double-bootstrapping technique was firstly introduced, then two set of fishing data involved two different fishing species were used as case studies to show and compare how this technique was used to estimate the uncertainties of codend size selectivity. Our results demonstrated that the uncertainties of codend size selectivity could be easily estimated applying the double-bootstrapping technique. Additionally, the most commonly wrong practice of using the doublebootstrapping technique was discussed. Our study will provide a guideline for conducting trawl codend selectivity studies, improve the level of selectivity researches, and contribute to the sustainable development of marine fisheries in China.