This paper presents a design procedure for integrating reactor networks when several multicomponent feeds exist. The problem has been solved in two stages. The reactor network structure in which the most comprehensive product distributions could be attained has been created in the first stage. The generated reactor structure has provided means for achieving all attainable concentrations and yields from the given reaction scheme. In this stage no assumptions on the objective function and process structure have been made. Once such a network structure has been set the optimal reactor network has been extracted from the anticipated structure by optimizing objective function which can involve either cost estimates or the desired product distribution. The proposed procedure is illustrated by an example.