The kernel estimator especially the univariate type often needs one smoothing parameter as against more smoothing parameters demanded by greater dimensional estimators though it all depends on kind of smoothing parameterizations employed. The smoothing parameter(s) of kernels with higher dimension may be called smoothing matrices. Kernels of higher dimensions have three kinds of parameterizations as estimators viz: constant, diagonal and full parameterizations. Unlike the full parameterization, the diagonal parameterization exhibits some levels of restrictions. This study investigates the efficiency of kernel estimators to which smoothing parameterizations are applied. The asymptotic mean-integrated squared error is employed as a criterion function with emphasis on bivariate case only. With real data, the results show that full smoothing parameterization outperformed the constant and diagonal parameterizations in respect of the asymptotic mean-integrated squared error’s value and the kernel estimate’s ability to retain the true characteristics of the affected distribution