This paper aims to test several hypotheses on the determinants of the quality of trade in cross-country regressions, taking a sample of trade competitors in EU markets. The hypotheses are those underlying two models of VIIT: the so-called neo-H-O model based on factor endowment and an ?economic geography? model based on market size and economic integration. As the explanatory variables used (proxies for human capital, physical capital, market size and a dummy for market integration) significantly affect the dependent variable (unit-value differences), it seems plausible to conclude that these variables give rise to specialisation in different segments of the quality spectrum. Much information is drawn from the analysis with respect to CEE specialisation in low-quality exports to EU markets. In particular, the estimates suggest the existence of a process of ?crowding out? of the existing human capital due to the process of economic transition. Moreover, the smaller market size of the EU accession countries could contribute to strengthen the disadvantage in high quality segments of production. In fact, the significant coefficient of the variable used to measure market size suggests that liberalisation might be accompanied by increased concentration of high-quality productions in large markets. However, the geographic proximity to the core of Europe could counterbalance this force. The integration process itself could accelerate the process of catching up in terms of quality of products and of per capita income, providing Eastern producers with a larger market and potential for economies to scale.