Traumatic brain injury (TBI) is the leading cause of death in alpine skiing. It has been found that helmet use can reduce the incidence of head injuries between 15% and 60%. However, knowledge on optimal helmet performance criteria in World Cup alpine skiing is currently limited owing to the lack of biomechanical data from real crash situations. Purpose: This study aimed to estimate impact velocities in a severe TBI case in World Cup alpine skiing. Methods: Video sequences from a TBI case in World Cup alpine skiing were analyzed using a model-based image matching technique. Video sequences from four camera views were obtained in full high-definition (1080p) format. A three-dimensional model of the course was built based on accurate measurements of piste landmarks and matched to the background video footage using the animation software Poser 4. A trunk-neck-head model was used for tracking the skier's trajectory. Results: Immediately before head impact, the downward velocity component was estimated to be 8 m.s(-1). After impact, the upward velocity was 3 m.s(-1), whereas the velocity parallel to the slope surface was reduced from 33 m.s(-1) to 22 m.s(-1). The frontal plane angular velocity of the head changed from 80 radIsj1 left tilt immediately before impact to 20 rad.s(-1) right tilt immediately after impact. Conclusions: A unique combination of high-definition video footage and accurate measurements of landmarks in the slope made possible a high-quality analysis of head impact velocity in a severe TBI case. The estimates can provide crucial information on how to prevent TBI through helmet performance criteria and design.