In this paper, we introduce a new, computationally attractive estimator of long memory by taking a weighted average of the GPH or local Whittle estimator over different bandwidths. We show that the new estimator can be designed to have the same asymptotic bias properties as the bias-reduced estimators of Andrews and Guggenberger (2003) or Andrews and Sun (2004) but its asymptotic variance is smaller than that of the latter estimators. We establish the asymptotic bias, variance, and mean-squared error of the weighted estimators, and show their asymptotic normality. Furthermore, we introduce a data-dependent adaptive procedure for selecting r, the number of bias terms to be eliminated, and the bandwidth m and show that up to a logarithmic factor, the resulting adaptive weighted estimator achieves the optimal rate of convergence. A Monte-Carlo study shows that the adaptive weighted estimator compares very favorably to several other adaptive estimators.