The selectivities of 30 mm diamond mesh codend (30D) and square mesh codends with mesh sizes of 20 mm (20S), 30 mm (30S), and 40 mm (40S) were studied using the parallel haul method and a logistic curve on the coast of Zhaitang Island, Qingdao, in spring 2014. Diamond codend of 16 mm (2Multivariate statistics and the maximum likelihood method indicated that all experimental codends were effective in releasing of some economic fish species relative to the control net. Square mesh codends showed better selectivity than diamond mesh codends. The for 30S were 4.61% and 16.80% higher than those for 30D, and the 50% retention body length ( 38.47% and 61.29%Comparison with catch data using square mesh codends showed that the was 6.58 cm and 7.20 cm, and the selection rage (SR) was 3.02 cm and 0.87 cmof was 2.88 cm, 4.03 cm and 10.49 cm, and the SR was 0.74 cm, 0.52 cm and 2.67 cm for the 20S, 30S and 40S, respectively. Thus, t increased with larger square codend mesh size. A comprehensive analysis of the effects of different mesh types on the economic value of the catch, the release ratio of low economic value captures, and labor input of sorting catches all suggest that the use of 30 mm square mesh codend the fishery by reducing unwanted bycatch and therefore should be implemented in the region.