In conventional IP networks, packets are routed along the shortest path based on a static link-cost metric. This shortest-path routing approach such as in RIP, EIGRP, and OSPF works well; however, under heavy loads, it can lead to long delays and unbalanced traffic distribution, which creates congestion hot spots in the network where some links remain underutilized. These problems occur because bandwidth availability and traffic characteristics are not considered in routing decisions. The genetic algorithm technique is a prominent technique that can be used to solve the optimization problems. Its appeal is due to the fact that it is computationally simple, yet provides reliable and sufficiently accurate solutions with powerful parallel search capabilities. Since a routing problem in a computer network can be classified as an optimization problem; therefore, this dissertation proposes a traffic engineering solution using genetic algorithms (GAs) to minimize communication latency and provide an optimal load balancing. By combining advantages of the MPLS technology and RSVP-TE protocol, the Genetic Algorithm based Traffic Engineering can be seamlessly integrated into the Internet Protocol