A mobile ad hoc network (MANET) is a collection of wireless stations that are dynamically and arbitrarily located in such a manner that the interconnections between stations may change on a continual basis. To facilitate communication among the active stations on the network, a routing protocol is used to discover routes between stations. A routing protocol plays an important role for the overall performance of MANETs. A variety of routing protocols for MANETs have been developed by network researchers and designers primarily to improve the performance of MANETs with respect to correct and efficient route establishment between a pair of stations for message delivery. Examples of commonly used MANET routing protocols include optimized link state routing protocol (OLSR), ad-hoc on-demand distance vector (AODV), dynamic source routing (DSR), and temporary ordered routing algorithm (TORA). A good understanding of the effect of each of these routing protocols on a typical IEEE 802.11 network will assist an efficient design and deployment of appropriate MANETs. This research aims to study the impact of routing protocols on MANETs by simulation experiment. In the investigation four routing protocols OLSR, AODV, DSR and TORA were used and the impact of network performance under varying network sizes, node mobility and traffic loads was considered. Experimental results show that TORA performs better under high traffic loads for medium and large sized networks. DSR is best suited for small and low mobility networks. The aggressive behaviour in flooding the network for valid routes may not be suitable for large network with high mobility. Despite its aggressive behaviour towards mobility DSR also performs well in large networks with high node mobility. AODV perform better for medium sized networks under high traffic loads. OLSR perform best in most environment, however, it suffers and degrades when mobility and traffic load increases. This dissertation discusses the performance evaluation and comparison of four MANET routing protocols in different simulation scenarios drawing valuable conclusions and future improvements.