This talk presents work by Sundar Vishwanathan. Imagine you are going to build $p$ service centers and wish to know the best set of locations for them. Examples where this problem arises include fire stations, and ambulance centers. The $p$-center problem is to place the service centers so that no vertex is too far away from its nearest center, i.e., to minimize the maximum distance to the nearest center. This problem is known to be NP-complete. When distances are symmetric ($d_{ij}=d_{ji}$), we can achieve a 2-approximation algorithm. For the asymmetric case, no constant-factor approximation algorithm is known. In this talk, we present an algorithm for the asymmetric problem which achieves an $O(\log^* n)$ approximation factor.