||A "good" fit between the foot and footwear is important to prevent foot discomfort and/or injury. Many techniques are available to determine the overall shape of a foot. However, most of these techniques are unable to accurately map the anatomical landmarks of the foot. The literature has reported techniques using pattern recognition to obtain these critical points. However, their reliability is not well known. In this study, a reliable and accurate method to digitize the complete foot in three dimensions using a commercially available sonic digitizer is proposed. The points obtained can then be used for surface modeling thereafter. Sonic digitizers are not very popular due to their inherent variability during actual digitization. A set of different experiments were designed and conducted to evaluate and quantify these "error" variations. In the first experiment, the effective digitizer volume was evaluated and the tolerance of the equipment was validated. In the second experiment, error variations when digitizing soft surfaces were determined. The equipment bias was evaluated using mechanical gauge blocks. The operator effect was also determined when digitizing an actual foot. Based on these experiments, a methodology is proposed to digitize the human foot in three-dimensions with fairly good precision while obtaining important landmark information to better understand the human foot.