This paper provides a comparative analysis of SLAM algorithms applied at Formula Student AI (FS-AI) using data collected by the University of Glasgow’s FS-AI team, UGRacing, at the 2023 competition. As part of the international autonomous driving competition, UGRacing have developed a software capable of racing a Formula Student car through unknown circuits. An essential part of the software of UGRacing and other FS-AI teams across the globe is SLAM, which allows the car to memorise track boundary positions and estimate the car’s position on the track. Optimising this component enables a vehicle to calculate and follow an ideal racing line, leading to significantly improved lap times. Hence, this paper examines the performance of three state-of-the-art SLAM approaches: EKF-SLAM, FastSLAM and GraphSLAM. Experiments suggest that EKF-SLAM outperforms FastSLAM and GraphSLAM in terms of car position and landmark position estimates and computational efficiency. The dataset is publicly available at https://gitlab.com/ugrdv/fs-ai-datasets .