The recent fast advances in wireless mobile communication technologies have led to a wider range of requirements. Currently, third generation (3G) wireless cellular networks aim to provide multimedia services in addition to voice and data. However, the challenges facing 3G and next generation cellular networks are in maintaining the quality of service (QoS) requirements for each class of the multi service traffic in face of the limited radio resources and the mobility of users. In addition, the growing demand of mobile users for integration with the global communication infrastructure would necessitate their demand for having reliable services similar to wired communication and data networks. Accordingly, many issues should be addressed in 3G and next generation cellular networks to achieve the requirements for seamless roaming and internetworking. The aim of this project is to develop an adaptive scheme for Call Admission Control (CAC) for multi-class service wireless cellular networks. The proposed CAC scheme is based on complete sharing of the available bandwidth among all traffic classes. This work proposes a CAC adaptive scheme that can be achieved through call bandwidth borrowing and call preemption techniques according to the priorities of the traffic classes. However, maintaining the QoS in each class to avoid performance deterioration is essential. Therefore, this research introduces mechanisms for call bandwidth degradation, and mechanisms for call bandwidth upgrading based on the Min-Max and Max-Min policy for resource deallocation and allocation, respectively. In addition, the scheme applies call preemption according to the class of calls priorities, and call reactivation whenever the traffic conditions permit. In general, the proposed adaptive CAC scheme assumes an online monitoring approach of the system, to determine the amount of bandwidth to be borrowed from calls, or the amount of bandwidth to be returned to calls. The proposed adaptive CAC scheme is evaluated by simulation for performance evaluation. A discrete event simulation model is developed for the scheme, using randomly generated calls, assuming a one cell in a wireless cellular network. The simulation results show that the proposed CAC scheme using bandwidth adaptation techniques improves blocking probability, dropping probability, and bandwidth utilization significantly. In addition, the simulation results show that some parameters can improve the performance of the simulation such as the utilization threshold.