A number of important control theory problems are involved in the selection of input-output schemes of a given process and one family of such problems is referred to as Model Projection Problems (MPP); these problems deal with the selection of effective sets of inputs, outputs out of larger potential sets of inputs, output respectively. The aim of this paper is to classify the different types of MPPs, discuss their relevance is the Early Process Design Stages, provide a framework for their study and suggest tools, methodologies, which are essential for their solution. The dominant idea running through the present treatment of MPPs is that the suggested solutions aim a producing final models with inherently “good” control structure characteristics.