Abstract As medical data acquisition has become digital and tomographical scanners allow three-dimensional reconstruction in ease, it is still standard to analyze the data plane by plane or in a multiplanar view. Both are limited to two-dimensional environments and cause difficulties in analyzing complex anatomical and pathological structures. Decreasing hardware prices and PC-clusters, which can compare with more expensive super-computers, will help implementing virtual environments in medicine for teaching, simulation and interacting with medical image data. Using free open source software, such as 3D-Slicer, for segmentation and preparing the data for the CAVE is quick and suitable for clinical use as well as for education. For educational purposes, the use of very high-resolution data sets will be ideal and could be made available for everyone, based on the open source philosophy.