This study adopted a directed network to model multi-destination tourism. We analyzed the network model using the 2018 Visitor Profile Survey of the Hong Kong Tourism Board (HKTB) which incorporated 23 countries or regions. We found that the network of the destinations has a high density, clustering coefficient, as well as a short diameter. As the network becomes sparser, the degree of nodes appears to follow a power-law distribution. We also found that nodes with high in-degree centralities also have high closeness centralities,while the betweenness centralities of nodes are very low in the network. The strongly connected components suggest that some Western European countries, the U.S. and Australia have high connectivity in both directions. This suggests that these countries are both the key source markets and destinations of themselves in the network.