This article proposes a combined theoretical and experimental approach to assess and quantify the global uncertainty of a high-speed camera velocity measurement. The study is divided in five sections: firstly, different sources of measurement uncertainties performed by a high-speed camera are identified and quantified. They consist of geometrical uncertainties, pixel discretisation uncertainties or optical uncertainties. Secondly, a global uncertainty factor, taking into account the previously identified sources of uncertainties, is computed. Thirdly, a sensibility study of the camera set-up parameters is performed, allowing the experimenter to optimize these parameters in order to minimize the final uncertainties. Fourthly, the theoretical computed uncertainty is compared with experimental measurements. Good concordance has been found. Finally, the velocity measurement uncertainty study is extended to continuous displacement measurements as a function of time. The purpose of this article is to propose all the mathematical tools necessary to quantify the individual and global uncertainties, to highlight the important aspects of the experimental set-up, and to give recommendations on how to improve a specific set-up in order to minimize the global uncertainty. Taking all these into account, it has been shown that highly dynamic phenomena such as a ballistic phenomenon can be measured using a high-speed camera with a global uncertainty of less than 2%.