Demand responsive transit (DRT) services are expensive to operate, so it is crucial to understand the factors that drive the agency costs and be able to investigate and predict the effect of possible changes in specific operating parameters as well as demand and service coverage areas on total costs. In this research, we adopt the concept of the analytical model proposed by Daganzo (1978) to present continuum approximation (CA) models that can estimate important service related factors. As the first step, we propose three analytical models to approximate the operation related variables of DRT services; Fleet Size, Vehicle Hours Traveled (VHT), and vehicle Miles Traveled (VMT). Since these variables implicitly control the operating cost of DRT services, an agency cost model is presented as a function of the operation models. This is a new approach for evaluating the operation and cost of DRT systems since the existing research addresses this issue with simulation models. Moreover, we present analytical formulations to explain the relation between zoning and no-zoning strategies on DRT system’s performance and agency costs. vi Different geometric structures to split a service area into two zones and the total agency cost formulation of each structure are investigated in this study. We identify the locations where zoning strategies can reduce the total agency costs and how zoning should be done in order to be most cost effective. Since demand is affected by the quality of service, especially travel time, splitting an area and introducing transfers will have an additional effect on changing the relative travel times between certain origin-destination pairs. Therefore, we propose a travel time model and a trip distribution model and link these models with our operations models in order to account for the interaction between zoning and trip distribution. The results of our models provide insights about how demand response trips are distributed within the service area and how the origin destination flow changes with zoning and transfer implementations. Finally, the models and methods used in this research are valuable to any agency operating DRT systems. It enhances their understanding of the system's performance, the relation between their total cost, coverage area, demand rate and their service characteristics, and more importantly provides them with innovative approaches to keep their agency costs low. So far, our models have resulted in providing toolkits for New Jersey Transit and Pioneer Valley Transit Authority that simply help them estimate the effect of different operating practices on their total costs, total vehicle hours traveled and total vehicle miles traveled for their provision of demand responsive service.