Abstract Flight operation control still can be improved in some aspects. For example, solve the problems of risk-related factors and complicated assessment methods in flight operation, reduce the potential risk caused by the difference of subjective judgment of personnel, and improve the airlines’ ability to control flight operating risks. In order to achieve these improvement, this study proposes an environmental risk assessment method for flight operation control based on support vector regression machine. The method includes the following steps: (1) firstly, the flight phases are divided; (2) secondly, the main environmental factors of each phase are determined and the data normalization rules are set according to business knowledge; (3) lastly, the model is trained with historical flight data from an airline company, to assess the environmental risk in take-off and landing phase during flight operation. The experiment results prove the validation and efficiency of the model which has been applied to flight operation risk assessment in practice.