Abstract: Against the background of global industrial chain restructuring, accelerating energy transition and rising resource security risks, cross-border flows of metal minerals have become increasingly networked and dynamic. Based on bilateral trade data for 101 countries from 2010 to 2024, this study constructs multi-period directed trade networks and applies a Stochastic Actor-Oriented Model（SAOM） to identify the drivers of the formation, persistence and adjustment of metal mineral trade ties. The results show that the global metal mineral trade network is characterized by structural selectivity: the density effect is significantly negative, while the reciprocity effect is significantly positive. Trade network status has positive sender and receiver effects, indicating that countries in core positions are more likely to expand export ties and attract import ties. The negative status similarity effect suggests that trade relations tend to form between countries at different network levels, reflecting complementary links between core countries and peripheral or semi-peripheral countries. Additional mechanism tests indicate that technological innovation and cross-regional market search capability are important channels through which trade network status promotes the expansion of metal mineral trade ties. Heterogeneity tests by mineral category, international cooperation network and development stage further reveal that the evolution mechanisms of trade relations differ across minerals, cooperation networks and country groups. This study extends research on global metal mineral trade networks from a dynamic network perspective and provides empirical evidence for optimizing international cooperation and differentiated resource security strategies.