The demand and capacity of a van dial-a-ride service for elderly and disabled passengers are analyzed together. The service demand is based on a survey of an advanced public transit system for an elderly and disabled persons van service. Demand is estimated by the Poisson regression on individual demand and by aggregating over individuals in the market. Service supply is modeled in two ways--a deterministic queuing model and a regression model--and is estimated by using drivers' manifest data. Market demand is compared with market supply derived from each method to identify the equilibrium level of ridership and travel time, which are used to measure the service quality. The model is used to predict a tradeoff between ridership and travel time and to present this tradeoff conveniently on a simple tabulation form. The service provider can use this form to determine combinations that maximize the management objective without engaging in simulation or estimation for all possible choices.