In the last decade, air traffic volume in East Asia has increased rapidly, and flights in the region suffer from severe delays. Nevertheless, most previous studies on flight delay determinants were conducted in North America and Europe, and very few of them were based on East Asia. Furthermore, previous research focused on such determinants from airport and route perspectives and disregarded the network perspective (i.e., influence of “connected airports”). This study aims to explore the determinants of flight delays in East Asia, and it covers not only airport and route determinants explored in previous research but also those regarding directly connected airports to include the network perspective. This study adopts panel data regression models to verify how the variables of the three perspectives influence flight delays in East Asia. The flight data of 4,611 routes among 318 airports in the second half of 2017 are used. The daily average arrival delay of routes is employed as the dependent variable, and three perspective and control variables are utilized as independent variables. The empirical findings indicate that from the airport perspective, congestion internalization and the “hubness” effect suggested in previous research do not exist in East Asia. The route variables affect delays in various countries and route types differently, and connected airports are crucial determinants of flight delays in East Asia. These findings imply that the experiences of and theories from North America and Europe are inapplicable in explaining flight delays in East Asia. This study not only addresses the gap in existing literature but also provides a meaningful basis for the establishment of airport congestion tax, aircraft deployment, and other operational carrier strategies for East Asia.