On the basis of the longline survey targeting albacore tuna (Thunnus alalunga) data collected from September 2012 to November 2012 in the waters near Cook Islands, the soak time calculation models of every branch line in each operation were developed by both models of hook retrieval and the hook depth prediction model of every branch line was also developed. The hook depth data were assigned to six layers of 40 m each (40-80 m, 80-120 m,120-160 m,160-200 m,200-240 m,240-280 m). The numbers of hooks and individual of fish in the whole water bin and each depth layer were estimated. The soak time of longline gear in the whole water bin and each depth layer were estimated. The soak time of longline gear were divided into one hour interval for the quantity of hooks and the individuals of albacore tuna, respectively. The respective catch rate of albacore tuna in each hour interval was calculated. The results showed that (1) the quadratic curves could be used to fit the relationship between soak time and the catch rate of albacore tuna, the catch rate of albacore tuna showed increasing at first and then decreasing trend with the increase of soak time; (2) the catch rate of albacore tuna was the highest when soak time was 11.0～11.4 h in the 40～280 m water body and six water layers. This study suggested that (1) the soak time of each hook lasted about 10.0～12.0 h in the longline operation targeting albacore tuna for improving the fishing efficiency; (2) for the pelagic longline, the optimum soak time of the whole water bin can be used to indicate the optimum soak time of various depth layers; (3) the optimum pelagic longline soak time varies with the target species; (4) the soak time of the longline gear could be considered as the effective fishing effort. The results will be applied to improve the fishing efficiency and will be used for the references to the fishing strategy and CPUE standardization.