The EUROP classification system is based on visual assessment of beef carcass conformation and fatness. The aim was to test the accuracy of the trained and certified abattoir EUROP classifier in Poland relative to national assessors. The results of EUROP system, carried on the 3135 carcasses were analyzed. Assessments were performed in the same conditions on the slaughter line by four independent classifiers. On the basis of the results of beef carcasses evaluations statistical analysis were performed. The results showed that the repeatability and accuracy of the national senior assessors was good. The results of evaluations abattoir assessor are different in comparison to three national assessors (p<0.01). The results show, that fat subclasses are the largest predictors of the differences in EUROP evaluations. The results suggest that visual assessment of beef carcasses might not be objective and is biased with error evaluator.