This paper presents a new control-loop configuration criterion for multivariable processes. For an arbitrary loop, four cases corresponding to different combinations of open and closed states with the remaining loops are investigated. A new interaction measurement, which is able to provide a comprehensive description of interaction among loops, is proposed to evaluate the loop-by-loop interaction. Consequently, a new loop-pairing criterion based on the new interaction measurement and the algorithm for determining loop pairings that result in minimum loop interactions in terms of interaction energy are proposed. The main contribution of this work is that it systematically analyzed the loop interaction from all aspects and derived a feasible solution for the problem. Several examples, for which the RGA-based loop-pairing criterion gives undesirable pairings, are provided to demonstrate that the proposed interaction measurement and variable-pairing criterion are simple, accurate, and easy to implement.