In this paper we empirically assess the evolution for the EU regions of both employment and unemployment before and after the Global Crisis. After a review of the literature on the theories and key determinants of regional unemployment, we shall overview the main findings concerning the labour market impact of the Global Crisis. The empirical analysis will initially be carried out at the national level including all EU countries; subsequently, we shall focus on the EU regions (at the NUTS-2 level), in order to detect possible changes in the dispersion of regional unemployment rates after the crisis. Our econometric investigations aim to assess the effect, on labour market performance, of previous developments in regional labour markets time series, as well as the importance of structural characteristics of the labour markets, in terms of the sectoral specialization of the regional economies. In fact, the local industry mix may have played a crucial role in shaping labour market performance in response to the crisis. In addition, we consider further characteristics of the regional labour markets, by including indicators of the level of precarization of labour and of the share of long-term unemployed, as indicators of the efficiency of the local labour markets. From a methodological viewpoint, we exploit eigenvector decomposition-based spatial filtering techniques, which allow us to greatly reduce unobserved variable bias -a significant problem in cross-sectional models -by including indicators of latent unobserved spatial patterns. Finally, we render a geographical description of the heterogeneity influence of past labour market performance over the crisis period, showing that the past performance has a differentiated impact on recent labour market developments.