Discussion of techniques for quantitative determination of the optimal prices and service quality in a wide class of systems. A probabilistic demand model sensitive to both price and quality is derived from the microeconomic concept that a consumer chooses the service that maximizes the difference between willingness to pay and price. This model is then aggregated to obtain a partial equilibrium macroeconomic model of demand. A Bayesian technique is developed for using data to reduce uncertainty about consumers' values. Following this, criteria are listed for optimizing prices and choosing among alternative qualities of service. A stochastic approximation is applied to finding optimal prices in service systems; in the cases tested, only a marginal increase in aggregate social welfare is obtained from establishing multiple priorities. A major shift occurs, however, in the incidence of benefits. 184 pp. Ref.