A mobile adhoc network is a collection of autonomous mobile nodes that that communicate with each other over wireless links. Hosts need to specify the requirements of the neighbors for efficient routing outside their transmission range. However, since there is no stationary infrastructure such as base stations, each host has to act as a router in itself and hence, a routing protocol runs on every host and is therefore subject to the limit of the resources at each mobile host. This paper proposes a new Efficient and dynamic probabilistic broadcasting (EDPB) approach, which solves the broadcast storm problem of AODV. Global Mobile Simulator is used to run simulations. The main performance metrics considered in simulations are routing overhead and end-to-end delays. The results show that, the normalized routing load is reduced to around 35% to 40% compared with AODV-blind flooding and AODV-fixed probability model, at a very heavy traffic load, when used with AODV-EDPB. The results show that AODV-EDPB algorithm performs better than AODV-blind flooding and AODV-fixed probability in terms of packet delivery ratio, end-to-end delay etc. especially in dense networks.