Crab pots have become one of the main fishing gears for harvesting swimming crabs in the East China Sea since the 1980s. However, the tremendous fishing effort of the crab pot fishery in the East China Sea has brought great pressure on crab resources. Installation of escape vents on crab pots is one of the potential technical measures to improve the fishing efficiency of pots and conserve crab resources. In this study, to examine the effect of the location of escape vents on escaping behavior of Portunus trituberculatus in crab pots, the response behavior of P. trituberculatus to three rectangular escape vent configurations, including B32 vent, escape vent with 3.2 cm opening height and located at the bottom edge of the side panel, B40 vent, escape vent with 4.0 cm opening height and located at the bottom edge of the side panel, and T40 vent, escape vent with 4.0 cm opening height and located at the top edge of the side panel, was observed in a tank by infrared underwater cameras. Generalized linear mixed models (GLMM) with logit distribution were used to analyze the effect of the location of escape vents and quantitative behavioral indicators derived from underwater observations, including the number of crabs, duration of stay in pots, and so on, on the escaping behavior of crab. The results from underwater observation indicated that P. trituberculatus pre-stocked in the pots could detect and approach the bottom vents readily in their search behavior on the floor of pots. The detection and approaching of the top vents only took place in their vertical search behavior. Although the durations from ingress to first approaching and mean durations of all approaching the top vents ware more than those for bottom vents, no significant difference in the mean number of times of approach was detected between the top and bottom vents. For crabs approaching vents, 87% of them attempted to pass through the vents and there was no significant difference in the attempt to pass through between bottom and top vents. The model fit of GLMM indicated that no significant quantitative behavioral indicators affected the escape attempts of crabs. All crabs walked sideways through vents, so whether crabs pass through an escape vent depends upon carapace depth. To better match the top escape vents, crabs spent more time and effort adjusting their posture and location. However, the output of GLMM fit shows that the location of escape vents has no significant effect on the ratio of crab escape, which indicates that P. trituberculatus are good at manipulating their body, while the effect of the number of crabs inside pots was a significant factor. According to behavior observation, it is suggested that the efficiency of the escape vent could be affected by the activity of search behavior of crabs and soaking time. Taking into account the structure of the crab pot and the current status of fisheries, the authors recommended an escape vent mounted at the bottom edge of the side panel should be a feasible technical measure.