A new estimator called s-K-B was developed to handle multicollinearity in mixed effect linear models. The s-K-B estimators were found to be better than Ridge, Stein, s-K, and least squared estimators under specific conditions.