<p>This paper addresses the multifunctional nature of interconnected spatial systems shaped by a network of adjacent regions. The space-time dynamics of the socio-economic patterns of regions are usually co-determined by spatial and temporal dependences, so that also the resilience trajectories of regions are influenced by spatio-temporal autocorrelations. This study employs a new form of spatio-temporal principal component analysis of the multi-faceted socio-economic profile of spatially dependent regions. It seeks to integrate the global Getis–Ord spatial association index in a multifunctional and multivariate approach for the analysis of space-time dependence patterns of a multiregional system, in which both temporal and spatial variability of socio-economic indicators are incorporated. Another novel element of the study is that both positive and negative spatial autocorrelation phenomena are taken into account. Negative spatial autocorrelation suggests essentially a case of spatial vulnerability. This approach is empirically tested by means of an evidence-based examination of the temporal evolution – and resilience patterns – of distinct socio-economic indicators for 16 interconnected regions in Poland. The socio-economic and geographic mosaic of our spatially dependent network analysis is described by means of choropleth maps, which confirm the presence of both positive and negative spatio-temporal autocorrelation among several core socio-economic variables in the Polish space-economy. The analytical approach based on the global Getis–Ord index G enriches the tradition of quantitative research into mutually interacting socio-economic characteristics of complex spatial systems.</p>