This thesis is an exploratory and seminal work aimed to select and test available calibration algorithms for efficiency, usability and most importantly, accuracy of multiple camera and lens configurations in the context of DAS (Driving Assistance Systems). The camera and lens configurations considered were basically fixed stereo setups with normal and fish-eye lenses combined with low and high camera sensor resolutions. The stereo camera setup used for calibration and experiments were similar to those used during DAS experiments. The selected calibration algorithms were four: OpenCV calibration, Bouguet, Mei and Scaramuzza algorithms. The OpenCV calibration was selected and tested for normal lens while Bouguet, Mei and Sacaramuzza algorithms were selected for fish-eye lens. The methodologies selected and used for testing and comparing calibrations were backprojection error and row misalignment error as well as direct comparison of calibration parameters whenever applicable. The calibration experiment results showed that OpenCV calibration is a suitable and accurate calibration algorithm for normal lens in the context of DAS. Similarly,Bouguet's fish-eye calibration toolbox seems to be the most appropriate in terms of accuracy and robustness in the context of DAS according these calibrations experiments. Mei's algorithm was second and Scaramuzza was third mostly due to inaccuracy and difficulty to use. Finally, this research contributed to the utilization of multiple camera calibration in DAS systems as well as to the evaluation and recommendations for best camera configurations for different purposes and environment conditions.