In a mobile network, nodes are mobile resulting in a dynamic network topology. An increase or decrease in speed or change in direction may change the pattern of the mobility in a mobile network, which can strongly impact the network efficiency. In order to understand the impact of the mobility pattern it is essential to estimate the mobility patterns of the mobile nodes which can be achieved with the help of particular mobility models. Mobility patterns in a mobile network may affect the performance of routing protocols, as a change in mobility pattern may lose the best existing available route due to link failures. Multipath routing provides fault tolerance against such link failures. Distributed Hash Tables (DHTs) based multipath routing known as Multipath Dynamic Address Routing (M-DART) is considered one of the multipath routing protocols considered robust under mobile networks by distributing the node’s location information throughout the network. Mobility pattern in a network may adversely impact the performance of M-DART in different network scenarios. Thus, it is important to study the impact of various mobility models on the performance of M-DART. In this paper, we have analyzed the impact of a range of mobility models on the performance of M-DART with the help of extensive experiments in different network scenarios. The results of our analysis show that M-DART has better performance under group-based mobility models compared to entity-based mobility models.