A mobile ad hoc network (MANET) is a network with no infrastructure that has the following characteristics: node mobility, frequent topology changes, a lack of central coordination, and thus, On a MANET, each node serves as both a host and a router, or as a relay between the source and destination. Providing quality service applications in MANET environments is a more challenging task due to the nature of MANETs. AOMDV routing is a typical MANET routing protocol that uses “best effort” delivery techniques without any guarantee of QoS. So, we proposed to modify the AOMDV protocol to improve QoS by using available bandwidth and buffer size to select the best path in addition to hop count in the origina l AOMDV routing protocol. Each node calculates its available bandwidth by listening to its channel, and then encapsulates the route request (RREQ) packet during the route discovery process. Route replay (RREP) messages back to different paths from which they were received in the proposed routing protocol (MDAOMDV). We implement our modified MDAOMDV routing protocol on NS2, especially the NS2.35 simulator tool. We have evaluated the performance of modified MDAOMDV and the existing AOMDV routing protocol, following three output metrics: throughput, packet delivery ratio, and end to end delay. Finally, as demonstrated by the comparison of the original AOMDV protocol with the new modified MDAOMDV routing protocols, the modified protocol outperforms the original AOMDV protocol. Keywords: MANET, QoS, MDAOMDV, available bandwidth and buffer size