Abstract: Being one of the primary fishing targets in Chinese pelagic fishing, Pacific saury (Cololabis saira) is distributed in the subtropical to temperate waters of the northwest Pacific Ocean. In order to explore its resource status, according to the catch at size and biological data of Northwest Pacific saury from 2014 to 2018, we conducted a performance test and a sensitivity analysis on length-based cohort analysis (LCA) model and biomass-based length-cohort analysis (B-LCA) model. Besides, we applied Monte Carlo method to estimate the model parameters, resource quantity, fishing mortality coefficients and maximum sustainable yield of Pacific saury. The results show that: 1) The LCA model and B-LCA model exhibited excellent fitting abilities at 5, 10 and 15 mm length intervals, with stronger fitting abilities at 5 mm interval. 2) LCA model performed better for fishery data in units of number, while B-LCA model performed better for fishery data in units of mass. 3) Both LCA and B-LCA models were sensitive to changes in growth factor ( b ) and asymptote length ( L_\infty ), with higher sensitivity to b. 4) The average resource mass of Pacific saury from 2014 to 2018 estimated by LCA model was about 65.93×10⁴−171.51×10⁴ t; the fishing mortality coefficient was 0.529 2; the maximum sustainable yield was 37.73×10⁴ t. The average resource mass estimated by B-LCA model was about 47.88×10⁴−126.25×10⁴ t; the fishing mortality coefficient was 0.540 5; the maximum sustainable yield was 33.02\times 10^4 t. The maximum sustained production estimated by both models was lower than the average annual production of NPFC (North Pacific Ocean Commission) member countries ( 40.98\times 10^4 t), indicating that the Pacific saury resources had been overfished from 2014 to 2018.