Systematically organizing and analyzing the energy consumption structure of the world can reveal the organic connections between countries. Furthermore, such tasks can provide a global reference system for each country, enabling each one to adjust and optimize its energy consumption structure. Most previous studies of the global energy consumption structure overlooked the associations between countries and the evolutionary trends associated with the energy consumption structures of countries. This paper analyzes the evolution of the global energy consumption structure using an evolution tree model. The visual structure of this model provides a novel perspective for understanding and analyzing the underlying trends. First, 144 countries and regions are categorized into four different types using the k-means clustering algorithm. Countries and regions that belong to the same type generally follow similar evolutionary paths. Moreover, type IV countries, which are mainly developed countries, have the most diverse energy consumption structures. By contrast, the energy consumption structures of other types of countries and regions can be improved. Countries can be located in the global energy consumption structure of the evolution tree, and these locations can provide a basis for improving the energy consumption structure of a country based on similar countries that are more diverse. (C) 2018 Elsevier Ltd. All rights reserved.