Multipath routing allows building and use of multiple paths for routing between a source-destination pair. It exploits the resource redundancy and diversity in the underlying network to provide benefits such as fault tolerance, load balancing, bandwidth aggregation, and improvement in QOS metrics such s delay. There are numerous multipath routing protocol proposed for wireless and hoc networks exploring characteristics in mobility, interference, topology, etc. This paper provides an overview of the most recent trends in multipath routing protocols in MANET. Recently biologically inspired techniques attracted much attention and it is found that they have promising potential applications in MANETs since they are very suitable to build distributed, robust, adaptive, scalable, and environmentally aware systems. Biological inspired techniques include artificial neural networks, evolutionary computation, swarm intelligence, etc. When these techniques are used to design multipath routing, they give optimal set of routing paths in MANET, which is considered as NP-hard problem.