To promote the development of Antarctic krill ( Dana) fishery, we proposed two designs of the 4-panel trawl (dh256-1 and dh256-2). The net mouth perimeter of the two trawls was the same, and the width ratio of belly panel to side panel (WRBS) was different:1.00 for dh256-1 and 1.10 for dh256-2. Model experiments were used to study trawl performance, and to determine the optimized nets and accessories, which include the sum of buoyance and sinker weight (SBSW), bridle length, and horizontal opening coefficient ( was between 0.5 and 0.6 or when the trawl speed was under 1.5 m/s and the <0.05) was noted in the vertical net opening between the two trawls:the vertical net opening of dh256-2 was remarkably faster than that for dh256-1. For improving the trawl performance, the SBSW should be larger for dh256-1:the minimum SBSW for dh256-1 was 4.59 tf and that for dh256-2 was 1.98 tf. These differences might be related to the flow attack angle of the nets. The WRBS of dh256-2 was larger than that of dh256-1; the flow attack angle and head surface of belly panels increased and improved the vertical net opening. Con-versely, when the WRBS was smaller, the flow attack angle and head surface of side panels increased, and the pull of side line to bridle was strengthened, thereby limiting the vertical net opening. Therefore, when the WRBS is smaller, the SBSW should be larger. No significant difference in coefficient of the energy consumption ( was inversely related to the _e; when the bridle length was ≥ 50 m, the effect of bridle length to both the trawls was not obvious. Therefore, 50 m bridle needs to be used for the two trawls. The effect of SBSW on the of dh256-1 was related to the trawl speed:when the speed was ≤ 1.5 m/s, the effect of SBSW on C_e declined remarkably with an increase in SBSW. Since the trawl speed of the Antarctic krill is usually around 1.5 m/s, the dh256-2 was recommended for Antarctic krill fishery. These results can provide a basis for designing Antarctic krill trawls.