A closed processor-sharing (PS) system with multiple customer classes is considered. The system consists of one infinite server (IS) station and one PS station. For a system with a large number of customers, a saturated PS station, and an arbitrary number of customer classes, asymptotic approximations to the stationary distribution of the total number of customers at the PS station are derived. The asymptotics for the probability mass function is described by a quasipotential function, which defines the exponential decay for the distribution, and a state-dependent preexponential factor. Both functions have an explicit expression in terms of the solution at each point x of a polynomial equation whose order equals the number of classes and whose coefficients are explicit functions of x. The quasi-potential function at its minimum point provides the logarithmic asymptotics for the normalization constant, and the asymptotic approximation for the variance is inversely proportional to the second derivative of the quasi-potential function at its minimum point. The complementary probability distribution is computed using the normal approximation and its refinements, which do not require repeated solution of polynomial equations. Numerical results demonstrate the range of applicability of the approximations. The results can be applied to the problem of dimensioning bandwidth and of admission control for different data sources in packetswitched communication networks.1. Introduction. This paper is motivated by a new application of closed queueing networks (CQN) with a large number of customers. The application is the dimensioning of bandwidth and of admission control for different data sources subject to feedback control in packet-switched communication networks when available bandwidth at the network nodes is shared between all active sources. Data sources are modeled by an infinte server (IS) station, network nodes are modeled by processorsharing (PS) stations, and a "customer" in the PS station represents an active data source. We consider a CQN that consists of one IS station with multiple customer classes and one PS station. The distinguishing property of the new application is that this CQN model is valid only if the PS station is saturated; see [3] for further details. The saturated station is defined asymptotically as the station, where the number of customers grows proportionally to the total number of customers in the network as the latter increases with service rates at the PS station.The application includes the performance metric that the bandwidth received by an active data source at a given network node is greater than a target value with probability 1 − α, where α is in the range of 0.001 to 0.1. As the network nodes of interest have a packet-based implementation of PS, the performance metric can be restated as the number of active data sessions at a network node (the total number of customers at the PS station in the CQN model) is less than a target value with the given probability 1 − α. As the above prob...