An overwhelmed number of vehicles have wrecked the current system of transportation due to rapid growth in population. Smart cities are the novel innovation that is inevitable to curb the problems of traffic jams, unorganized traffic, environmental pollution, and slow response rate to emergency situations. The intelligent transportation system (ITS) is an integral part of smart cities allowing communications and interaction among vehicles. An autonomous vehicle is the key element of ITS and the mass implementation of this emerging technology is the solution to traffic problems linked to the current transportation system. Autonomous vehicles lead to the need for efficient and reliable external vehicular communications particularly through vehicular ad hoc networks (VANET). However, utilizing a suitable routing protocol to provide stable routing and efficient performance for vehicular communications in autonomous vehicles is a key factor. Routing protocols are particularly important for establishing vehicular to vehicular and vehicular to infrastructure (V2X) communication, which is incredibly challenging due to the movement of nodes. The quality of inter-vehicular communications is widely affected by numerous factors such as routing protocols, traffic environment, and traffic density. This article presents a detailed evaluation of three commonly used protocols, i.e., Ad-hoc On-demand Distance Vector Routing (AODV), Dynamic Source Routing (DSR), and Destination-Sequenced Distance-Vector Routing (DSDV) under three different traffic environments. To investigate the performance of these routing protocols under diverse environments, simulations are extended further by using the varying density of vehicles. This study aims at finding the best routing protocol for efficient and reliable packet dissemination among vehicles under different scenarios.