Pigs and dairy cows are the two main industries of livestock production in the Czech Republic (CR) and in the European Community (EC) as a whole. Of the total annual per capita meat consumption in the CR (81 kg), in the EC (92 kg) and worldwide (42 kg) pig meat accounts for 54% (43.5 kg), 47% (42.7 kg) and 39% (16.4 kg), respectively. Longterm forecasts indicate that the production and consumption of pig meat in the EC will slowly grow by about 3% to 4% by 2013. In accordance with the common agricultural policy reform, which is in the process of realization, increased attention will be paid to food production. Consequently, issues associated with the identification, classification and quality of pig carcasses will be of increased importance.According to Kirton (1989) pig classification is the most objective of the various systems applied to pig, sheep and cattle carcasses, because it is the animal to which it is easiest to apply objective classification methods. One reason for this is that it has more uniform fat cover than other species.The conventional method of pig carcass classification (FOM, SKG II) was analysed by Matzke et al. (1986), the AutoFOM system based on ultrasonic measurements was described by Brøndum et al. (1998), Busk et al. (1999), and the accuracy of FOM instruments used in on-line pig carcass classification in the CR was detailed by Šprysl et al. (2007) and others.The accuracy of the EU reference dissection method for pig carcass classification in four dif- The highest correlation coefficients are between QC and LM (r = -0.920), LM and FT (-0.900) as well as QC and FT (0.828), the lowest between FT and MT (r = -0.084). Quality class as the dominant indicator is influenced mainly by LM, which explains from 77% to 89% of variability in the case of linear regression. Among the eight methods of pig carcass classification the FOM apparatus was used the most frequently (46.5% carcasses) followed by the ULTRA-FOM 300 apparatus (15.6%), another apparatus (13.2%) and by the IS-D-05 unit (9.8%). In the statistical models used all effects (differences) are statistically significant because of the large size of the data set. The results from the separate evaluation of each cross-classified effect are that EV has the largest influence and yearseason and methods have a smaller influence. The time trend (42 months) documents stable CW and MT, a slight increase in LM and improvement of QC. The estimated results indicate the successful introduction of pig carcass classification in the CR after accession to the EU.