The state of a living body is mainly characterized by the presence of a constant flow of chemical reactions, and therefore living beings are formed by several reactions, linked to each other in a kind of network. The theory of biochemical networks provides a mathematical and computational framework used to analyze and simulate reactions such networks. It is a building model, diagnosis and network analysis based on ordinary differential equations. Some chemical reactions networks are studied mainly for the complex dynamic behavior. Far from equilibrium, such nonlinear dynamical systems exhibit phenomena such as, multiple steady states and oscillations. Some chemical networks of biochemical interest were analyzed, such as calcium networks in the cilia of olfactory receptor neurons, the glycolytic pathway, oxidase-peroxidase and network granting a multistable dynamics in the fruit fly embryogenesis, Drosophila melanogaster, as well as networks theoretical. To study them, the method Stoichiometric Network Analysis, or SNA, provided a systematic approach to the dynamics of chemical mechanisms or other systems containing stoichiometry. First it was necessary to understand the biochemical systems that exhibit complex dynamics and the importance of such in maintaining metabolic and cellular communication in living creatures to then assess the effectiveness of the stoichiometric analysis of technical networks in the investigation of nonlinear phenomena. The graphical variant of the SNA method was used to create subnets that exhibit unusual behaviors and the consequent use of the same in modeling elementary easy to grasp networks. Finally, the methods have been shown effective in analyzing complex dynamic, moreover, it can verify the bifurcation analysis for some networks. In addition to the necessary and sufficient conditions for observing complex dynamics, the SNA by technique could even expose another biological properties of the models.