Abstract: Under the background of climate warming, the frequency and intensity of droughts are increasing. The regularity of drought occurrence and the complexity of its formation mechanism are becoming more prominent, which poses new challenges to the mechanism study on drought formation, the theory and method of drought prediction and changes in disaster risk. They also restrict improvements on current drought prediction, early warning, disaster prevention and control ability. In recent years, with the support of several national projects such as the National Key Basic Research and Development Program (973 Program), a series of new achievements have been made in the studies of formation mechanism, prediction theory and risk characteristics of drought disaster. This paper systematically summarizes recent progress in the following aspects: (1) Clarified the mechanism of Qinghai-Xizang plateau heating, sea temperature, summer monsoon, remote correlation and other factors on the formation of drought. (2) Found the sensitivity of crops at different growth stages and temporal drought scale to precipitation deficit. (3) Revealed new characteristics of drought disaster risk distribution and variation in typical regions under the background of warming, and constructed a new conceptual model of drought disaster risk. (4) Developed the integrated forecast system of seasonal and sub-seasonal drought in East Asian monsoon area. On the basis of summarizing the existing research results, future research on drought formation mechanism and disaster risk is proposed, and five key research directions are put forward: (1) influences of multi-factor linkage and multi-scale superposed effect on drought formation; (2) studies on climate models that systematically integrate human activities and decisions with their associated feedbacks; (3) effects of land-atmosphere coupling and atmospheric circulation on drought; (4) knowledge of the key process of drought disaster impact on food security and ecological security; (5) research on improving the accuracy of drought prediction under different climate scenarios in the future.