Deepwater oil and gas is becoming great concerns in terms of the energy development in China. The exploration and exploitation of the deepwater oil and gas has been recently promoted to the plan of building a powerful ocean country and energy safety. However, the harsh and complicated offshore environment as well as the technological conditions still exit during the exploitation process. In view of the main operation scene, the research on major accidents was systematically carried out including the risk assessment, risk evolution, consequence assessment, accident control and so on. Firstly, the risk evolution methodology during the drilling operation was proposed and the reliability assessment model for the critical barriers was constructed. In addition, with regard to the potential risk during the deepwater oil and gas exploitation process, the robust risk assessment approach and the digital control scheme were developed by integrating the machine learning algorithms and data-driven models with the consequence modeling considering multi-object-based parameters. Altogether, we review the above achievements and also give the future research directions in this area.