Increasing demand for passenger services in densely populated urban environments, are currently covered overwhelmingly by private vehicles. Their impact on CO2 emission, present a serious obstacle to the reduction objectives, in the Netherlands alone the target of 45% by 2030, for limiting the global warming to 1.5°C degrees. Autonomous Vehicles (AV) and Ride-Sharing services are believed to be offering a crucial technological and perception shifts to reducing emission. In this work, a methodology for assessing the impact of a large-scale AV fleet ride-sharing system to replace the one-two passenger vehicle traffic using Rotterdam as the case study is designed and proposed. The approach includes three stages: 1. Building and finetuning a traffic model using publicly available data 2. Designing and implementing a trip merging component, in the form of two distinct heuristic greedy algorithms and a variation of the second one, using Python programming language. 3. Evaluating the impact of each merging scenario on the network in SUMO. The system’s influence and results are driven from the deployment of the ride-sharing service on the 2016 traffic model. The decrease in total number of trips, vehicle kilometres travels, and subsequent improvements in traffic flow resulted in 39% reduction in CO2 emission with the third algorithm. This result not only establishes the extent of AV ride-sharing service’s potential for emission reduction and traffic quality improvement. This adaptable methodology also operates as a proof of concept for a preliminary step for policy makers when considering implementing such service in any urban setting. Two of the major elements not included in this research are multimodal travel, like combination with public transport, and changes in demand for each mode choice based on traveller’s behaviour. These elements thus remain open for future consideration.