Abstract: To determine the suitability of minimum mesh size standards for trawls in the South China Sea, we conducted selectivity tests in that area by using the covered net method with mesh sizes of 30, 35, 40 and 45 mm (Denoted as D30, D35, D40 and D45, respectively). We used Logistic and Richards models to fit selectivity curves for Decapterus maruadsi by the maximum likelihood method, and determined the best-fitting curves by comparing the model deviations and degrees of freedom of the two models. The results show that the Logistic curve fitted better for D30, D35 and D45 mesh codend, while the Richards curve fit better for D40 mesh codend. The 50% retention fork lengths (L_0.5) of D. maruadsi for D30, D35, D40 and D45 were 12.43, 12.88, 13.60 and 16.79 cm, respectively. As the mesh size increased, the L_0.5 of D. maruadsi increased but the selection range decreased gradually. The minimum mesh size was 41.39 mm based on the relationship between L_0.5 and mesh size of D. maruadsi. Thus, 45.39 mm is suggested as the minimum mesh size for trawl in the South China Sea to protect small and medium-sized economic fish such as D. maruadsi.