It is of practical interest to identify which processes will benefit significantly from the use of constrained control algorithms such as model predictive control, and which will not. Explicit conditions are derived that identify whether a particular process may benefit from constraint handling. These conditions are also useful for understanding the interactions between design and control for a particular system, especially for actuator placement and selection. The conditions are computable for a large-scale system directly from its transfer function model, a simulation model (e.g. defined by a set of ordinary/partial-differential equations and algebraic conditions), or experimental input–output data. The formulation considers the effects of measurement noise, process disturbances, model uncertainties, plant directionality and the quantity of experimental data. The conditions are illustrated by application to a paper-machine model constructed from industrial data.