Rigorous analytical methods for consistent modeling of control structures for multivariable processes are presented. The continuous distillation process is used to illustrate the methods, but the basic ideas, and more specifically the transformations and consistency relations derived, are valid in general.If steady-state operating data and the process gains of an arbitrary control structure are known, it is possible to calculate the process gains of any feasible control structure. A general expression relating the process gains of different control structures is derived. In general, the process gains must also satisfy certain consistency relationships which can be derived from first principles, e.g., steady-state material balances.The usefulness of the results is illustrated by control structure transformations and reconciliation of process gains, by an application to process dynamics, by synthesis of noninteracting control loops, and by derivation of analytical relationships useful in relative gain analysis.