In this paper, a system of procedures for identification using the transient step response is presented. The identification aims at modeling dynamic processes with simple models that have general second-order dynamics. The procedures consist of three parts, including classification of the response selection of the model structure, and the estimation of parameters. Technically, this proposed system provides a strategy based on quantitative measures to identify the model structure and extends the identification of models to include possible RHP/LHP zeros and dead time. One important aspect of this proposed system of procedures is the minimization of the need for human decisions to be made during the identification process. With this advantage; it is then possible to facilitate the development. of software to perform the identification automatically.