We employ a spatio-temporal network modelling approach to identify possible relations between strong earthquakes and spatial regions worldwide. A global strong earthquake dataset containing 7736 events (Mw ≥ 6.0) from 1964 to 2018 is used. Statistical results identify power-law relationships and heavy tail phenomena in the spatial patterns of strong earthquakes. The interactions between regions follow the same law, with a few regions that may be hit by successive strong earthquakes with high probability. Also, we find that the interconnections between regions are mainly related to the succession of events in time, whereas the distribution of events is extremely inhomogeneous in space. This study provides a research prototype for the spatio-temporal analysis of global strong earthquakes, laying a foundation for obtaining insights into the network modelling approach for global strong earthquakes.