1: \begin{abstract} 

2: We reply to the comment by  K. Aryanpour, Th. Maier and M. 

3: Jarrell  (cond-mat/0301460) on our paper (Phys. Rev. B {\bf 65} 

4: 155112 (2002)).  We demonstrate using  general arguments and explicit 

5: examples that whenever the correlation length is finite, local 

6: observables converge exponentially fast in the cluster size, 

7: $L_{c}$, within Cellular Dynamical Mean Field Theory (CDMFT). 

8: This is a  faster rate of convergence  than  the $1/L_{c}^{2}$ 

9: behavior of the  Dynamical Cluster approximation (DCA) 

10: thus refuting the central assertion of their comment.

11:  

12: %The 

13: %very slow $1/L_{c}$ convergence of CDMFT, obtained by K. 

14: %Arayanpour et al., was due to a very inefficient way of 

15: %estimating local observables. 

16: \end{abstract}