The degree to which Advanced Public Transportation Systems (APTS) are expected to increase ridership is studied. The responses from a survey of Dial-a-Ride users to estimate the parameters of statistical models of user behavior were used. The respondents were asked to rank hypothetical scenarios in which level of service was varied. Level-of-service attributes included, among other variables, changes in travel time and ability to reserve the trip in advance. The respondents were asked to state how many trips they would have made during the past week of travel, given the hypothetical service attributes described in the question. The models developed link the average number of trips per week using Dial-a-Ride and the level of service offered by the Dial-a-Ride system. The modeling framework attempts to resolve the bias issues inherent in the use of stated preferences (SP) data, where respondents are asked to answer questions about a hypothetical situation. A basic methodology of combining revealed preferences and SP data is provided to evaluate the effect of different APTS configurations on ridership.