Air traffic controllers are the important parts of air traffic management system who are responsible for the safety and efficiency of the system.They make traffic management decisions based on information acquired from various sources.The understanding of their information seeking behaviors is still limited.We aim to identify controllers′ behavior through the examination of the correlations between controllers′eye movements and air traffic.Sixteen air traffic controllers were invited to participate real-time simulation experiments,during which the data of their eye ball movements and air traffic were recorded.Tweny-three air traffic complexity metrics and six eye movements metrics were calculated to examine their relationships.Two correlational methods,Pearson′s correlation and Spearman′s correlation,were tested between every eye-traffic pair of metrics.The results indicate that controllers′two kinds of information-seeking behaviors can be identified from their eye movements:Targets tracking,and confliction recognition.The study on controllers′ eye movements may contribute to the understanding of information-seeking mechanisms leading to the development of more intelligent automations in the future. Air traffic controllers are the important parts of air traffic management system who are responsible for the safety and efficiency of the system. The make make management decisions based on information acquired from various sources. The understanding of their information seeking behaviors is still limited. We aim to identify controllers’ behavior through the examination of the correlations between controllers’eye movements and air traffic. Sixteen air traffic controllers were invited to participate in real-time simulation experiments, during which the data of their eye ball movements and air traffic were recorded. Only -three air traffic complexity metrics and six eye movements metrics were calculated to examine their relationships. Two correlational methods, Pearson’s correlation and Spearman’s correlation, were tested between every eye-traffic pair of metrics. Results that results that controllers’two kinds of information-seeking behaviors can be identified from their eye movements: Targets tracking, and confliction recognition. The study on controllers’ eye movements may contribute to the understanding of information-seeking mechanisms leading to the development of more intelligent automations in the future.