Molecular network research is an important means to reveal the structure and function of biological systems from a global point of view. The existing analysis methods are mostly based on the static network. In fact, the biological networks change all the time with the environmental conditions, tissue types, disease status, growth and differentiation. Researchers have proposed some bioinformatics methods for the dynamic analysis of molecular networks, such as the dynamic classification of the hub nodes, the prediction of dynamic protein complexes, the construction of condition specific subnetworks, and the simulation of dynamic behavior of networks. In this paper, we reviewed these methods for the dynamic molecular network analysis. Predictable, dynamic network analysis will become the standard mode of the future network research.