Whenever the sign of the parameters are known of an econometric model, usual two-sided tests are no longer appropriate. In this situation, we propose a distance-based one-sided Wald test, a likelihood based test, to test one-sided alternative. Monte Carlo simulations are conducted to compare power properties of the proposed test with their respective two-sided counterparts. This text emphasizes the use of distance-based one-sided concepts in testing higher order autocorrelation in the context of linear as well as dynamic regression models.