For general metabolic networks whose reaction rates satisfy some conditions,a strong connectivity decomposition (SCD) method is proposed, which can not only divide the hole network into a set of sub-networks but also keep the strong connectivity of the network. his makes it possible to understand the multi-equilibrium property of the whole network by nalyzing the sub-networks. The SCD method is based on only the topological structure of he network. To get an SCD for a given metabolic network, the concepts of metabolic reaction raph, interaction graph and contraction operation are introduced. It is shown that for a given etabolic network, the SCD is unique, all the sub-networks are strongly connected, and the omputational complexity of the decomposition is polynomial. The relationship between the hole network and sub-networks is given in the sense of multi-equilibrium properties. Examples re given to demonstrate the effectiveness of the algorithms and the main results.