Taint analysis technique is the key technique means for analyzing the robustness of programs and vulnerability mining. By marking the data which are sensitive or untrusted, one can observe the flow of these tainted data during program execution, then determine whether the marked data affects the key nodes of the program. According to the implementation mechanism, the taint analysis can be divided into static taint analysis and dynamic taint analysis. As an auxiliary technique, it can be combined with mainstream vulnerability mining techniques such as fuzzing test and symbol execution, playing a great role in test case construction and path feasibility analysis. This article firstly introduces the basic concepts of dynamic taint analysis technique. Second, it focuses on the process of dynamic taint marking, propagation and detection. Then it summarizes the main defects in taint analysis and the application status of dynamic taint analysis technique. Finally, it is compared with the current mainstream taint analysis tools and explores the future trends of taint analysis technique.