Based on the measurements on 770 cagecultured large yellow croakers，the relations between fish body width, length and their weight have been analyzed by means of regression equation. From the geometric morphological measurements on fish body, which are often used in estimating gillnet selectivity and from Logistic curve model, the theoretic bar spacing of the grading panel has been derived as 34.66 mm to meet the requirement of the grading standard 300 g. It was 38.21 mm derived from a power function regression. Grading effects were tested with two different grading spacings of 33.01 mm and 37.95 mm. For the 33.01 mm grading spacing grading panel, after 5 to 10 minutes of grading operation, the highest ratio of marketable fish which weigh more than 300 g was 50.00% of total retained by the grading panel. Their average body weight retained by the grading panel was 296 g. For the 37.95 mm grading spacing in the same conditions, the highest ratio, whose weight ≥300 g was 73.33% in the total retained by the grading panel. The average ratio was 41.65%. The average body weight of total fish retained by the grading panel was 310 g which was 25 g more than that of fish without sizing grade. Grading trials have been operated in a 48 m circle fish cage. All of the fish retained by the grading panel, when the bar spacing was set at 37.95 mm, were heavier than 300 g and the lowest body weight was 320 g. The results show that grading effect was better when the bar spacing was 37.95 mm, which was derived by means of a power function regression equation.