An exploratory study of the value of ridge regression for interactive models is reported. Assuming that the linear terms in a simple interactive model are centered to eliminate non-essential multicollinearity, a variety of common models, representing both ordinal and disordinal interactions, are shown to have "orientations" that are favorable to ridge regression. Comparisons of the potential efficiency of ridge regression to that of ordinary least squares across a wide range of conditions (systematic variation of the degree of collinearity, sample size, strength of effect, and coefficient orientation) clearly suggest the value of ridge procedures for many centered interactive models. All calculations were conducted for population values, using the biasing parameters which are known to be "MSE optional" for generalized ridge regression (i.e., the results reflect the minimum MSE for ridge regression possible under the specified conditions). Four figures and one table are included. (Author/TJH)