In a previous paper (Peterson, Bertsimas, and Odoni 1992), we studied the phenomenon of transient congestion in landings at a hub airport and developed a recursive approach for computing moments of queue lengths and waiting times. In this paper we extend our approach to a network, developing two approximations based on the method used for the single hub. We present computational results for a simple 2-hub network and indicate the usefulness of the approach in analyzing the interaction between hubs. Although our motivation is drawn from air transportation, our method is applicable to all multi-class queuing networks where service capacity at a station may be modeled as a Markov or semi-Markov process. Our method represents a new approach for analyzing transient congestion phenomena in such networks. Airport congestion and delay have grown significantly over the last decade. By 1986 ground delays at domestic airports averaged 2000 hours per day, the equivalent of grounding the entire fleet of Delta Airlines at that tillie (250 aircraft) for one day (Donoghue 1986). In 1990, 21 airports in the U.S. exceeded 20, 000 hours of delay, with 12 more projected to exceed this total by 1997 (National Transportation Research Board 1991). This amounts to *School of Public and Environmental Affairs, Indiana University, Bloomington, Indiana tSloan School of Management, Massachusetts Institute of Technology, Cambridge, Massachusetts ;Department of Aeronautics and Astronautics, Massachusetts Institute of Technology, Cambridge, Massachusetts