Estimation of bycatch of the blue shark (Prionace glauca,BSH), a major bycatch shark in the Pacific tuna longline fishery, is a current focus of stock assessment and management. Based on the data collected during the 2012-2018 survey in the Pacific Ocean by national observers of tuna longline fisheries, the relationship between the bycatch of BSH and spatiotemporal, environmental, and operational variables was analyzed using the Tweedie-GAM model. The performances of the model-based approach and the stratified ratio estimation method on bycatch estimation were compared using cross-validation. The results showed that the number of hooks per basket (HBF), longitude, and latitude were the most influential variables, and the cumulative deviance explained by the optimal model was 51.6%. The bycatch fluctuated and decreased with the increase of HBF, and decreased significantly when HBF was more than 19. The spatial distribution of bycatch showed significant latitudinal and longitudinal increasing patterns, as it was lower and uniformly distributed in low latitude areas (10°S-5°N), and higher in eastern areas than in western areas. Regarding the temporal distribution, we found that the seasonal difference of bycatch was not significant, but the bycatch of fishing sets operating on a full moon day increased significantly due to the influence of moonlight brightness. Also, relatively high chlorophyll concentration (≥ 0.15mg/m³) or worm waters (≥ 29.5℃) was associated with high bycatch. There was no significant linear relationship between the bycatch and the number of hooks per set. The results demonstrated that the model-based bycatch estimation method was superior to the stratified ratio estimation method. To reduce the bycatch of BSH, it is recommended that the HBF of longline fisheries in the Western and Central Pacific should not be less than 20. The number of fishing vessels operating in the mid-latitudes and waters east of 178°E should be controlled, and their implementation of bycatch mitigation measures should be monitored in a focused manner. To improve the accuracy of estimating the total bycatch of BSH or similar stocks, subsequent studies are suggested to adopt model-based estimation methods.