Using a social network analysis, this study explored how the structure of international tourism has changed longitudinally. The international tourism remained relatively stable from 2002 to 2014. The networks of outbound and inbound international tourism have decentralized gradually. Based on a quadratic assignment procedure (QAP) analysis, each of four study years of international tourism network was found to be correlated. The groups formed by CONCOR analysis have similarities in terms of their continent, language, culture, economy, and population. Furthermore, outdegree, indegree, betweenness, eigenvector centrality are correlated with each other, and GDP is correlated with outdegree, indegree, betweenness, eigenvector centrality, and population. © 2020, Springer Nature B.V.