Summary: Under square loss, this paper constructs an empirical Bayes(EB) estimator for the parameter of a normal distribution which is both asypmptotically optimal and admissible. Moreover, the convergence rate of the EB estimator obtained is proved to be \(O(n^{-1})\).