In their Comment on our recent Letter [R.A. Klein, M.A. Zottola, Chem. Phys. Lett. 419 (2006) 254-258], Feller and Peterson point out that density functional theory combined with the Pople triple split-valence basis-set 6-311++G(2d,p), does indeed perform well in comparison to second-order perturbation and coupled cluster theory in combination with correlation-consistent basis-sets for the prediction of bond lengths and harmonic frequencies but does not provide acceptable accuracy for dissociation energies. MPW1PW91/6-311++G(2d,p) is, therefore, highly suitable and computationally efficient for generating starting structures for subsequent single-point (SP) calculations at higher and more computationally expensive levels of theory. (c) 2006 Elsevier B.V. All rights reserved.