Cross-regional free trade agreements (FTAs) have flourished in East Asia since the late 1990s. Japan and Korea were at the forefront of this trend. Nevertheless, most policymakers’ preferences orbited around region-oriented FTAs out of reluctance to make a sudden policy shift away from their conventional emphasis on World Trade Organization-based multilateral negotiations. The trend continued in Japan throughout the 2000s. In contrast, Korea took a sharp turn toward a cross-regional strategy in 2003. By the mid-2010s, this had created a significant gap in the two countries’ overall FTA partner choices. What caused policymakers’ ideas in the two countries to diverge from their initial focus on region-oriented FTAs? This paper focuses on the conditions that enabled policymakers’ ideas to explain the divergence by developing the contextual‒dynamic framework of FTA policies. At the dynamic level, individuals are more likely to emerge with new ideas when their expertise comes from outside the decision-making body, accompanied by their social qualities and power. Contextual conditions should also be met: the trade policy environment should reciprocate agent-level qualities to create a supportive atmosphere for policy change.