Urban sprawl is one of the main factors underpinning the changes observed in European landscapes and environment. The analysis of a phenomenon of this sort is conditioned in large part by the availability of documented urban databases, validated by experts and regularly updated. From this point of view the Urban Morphological Zones (UMZs) produced by the European Environment Agency form an important prospect for the future. However, it is essential to set the results of this highly automated approach against the already well-documented national databases. In this paper we investigate the validity of the UMZs for reflecting urban entities, through comparisons with Danish, French, and Swedish national databases for year 2000. We propose a method based on a matching protocol between UMZs and national agglomerations which is based on conceptual and technical choices, in order to take account of the complexity of urban objects. Two main results are pinpointed. On a global level the UMZ database is fully adequate for comparing levels of urbanisation among states. There is no systematic bias in the delineation of urban areas by the UMZs and total urban populations differ only by 5% to 8%. Conclusions should be more cautious when the scale of observation is the town or city itself. In France, mean differences are as small as in Denmark and in Sweden, but certain large agglomerations display marked divergences that depend partly on local urbanisation patterns. We propose a typology of deviations which could be useful for the majority of states that do not have a national morphological urban database, in particular in Eastern Europe.