The choice of medical imaging techniques, for the purpose of the present work aimed at studying the anatomy of the knee, derives from the increasing use of images in diagnostics, research and teaching, and the subsequent importance that these methods are gaining within the scientific community. Medical systems using virtual reality techniques also offer a good alternative to traditional methods, and are considered among the most important tools in the areas of research and teaching. In our work we have shown some possible uses of three-dimensional imaging for the study of the morphology of the normal human knee, and its clinical applications. We used the direct volume rendering technique, and created a data set of images and animations to allow us to visualize the single structures of the human knee in three dimensions. Direct volume rendering makes use of specific algorithms to transform conventional two-dimensional magnetic resonance imaging sets of slices into see-through volume data set images. It is a technique which does not require the construction of intermediate geometric representations, and has the advantage of allowing the visualization of a single image of the full data set, using semi-transparent mapping. Digital images of human structures, and in particular of the knee, offer important information about anatomical structures and their relationships, and are of great value in the planning of surgical procedures. On this basis we studied seven volunteers with an average age of 25 years, who underwent magnetic resonance imaging. After elaboration of the data through post-processing, we analysed the structure of the knee in detail. The aim of our investigation was the three-dimensional image, in order to comprehend better the interactions between anatomical structures. We believe that these results, applied to living subjects, widen the frontiers in the areas of teaching, diagnostics, therapy and scientific research.