Accuracy of application of USDA beef quality and yield grades using the traditional system and the proposed seven-grade yield grad system

Beef carcasses (n = 5,542) were evaluated by three USDA on-line graders and compared with the computed expert USDA quality (QG) and yield grades (YG) during 8-h shifts at a major beef-processing facility for a 2-wk period to evaluate the accuracy of applying USDA QG and YG within the traditional fiv e-grade and the proposed seven-grade (segregating YG 2 and 3 into YG 2A, 2B , 3A, and 3B) YG systems. Quality grade distribution of the carcasses was 1 .1% Prime, 50.0% Choice, 43.8% Select, and 5.1% No-Roll. Accuracy of applyi ng QG was not affected (P > .05) by changing from the five-grade (91.5%) to either the seven-grade system, when determining only QG (94.3%), or the se ven-grade system, when determining QG and YG (95.0%). Calculated expert YG successfully segregated carcasses into their respective YG, but on-line gra ders could not differentiate between YG 4 and 5 in the seven-grade systems. The application of YG in the five-grade system was more accurate (P < .05) than either of the seven-grade systems. A trend existed for on-line grader s to undergrade carcasses as the numerical YG increased. Total accuracy of applying YG decreased by 19.4 to 21.8% when switching from the five-grade t o the seven-grade system. The segmentation of USDA YG 2 and 3 into YG 2A, 2 B, 3A, and 3B resulted in a decrease in the ability of on-line graders to a ccurately apply the YG.