At present a rapidly expanding variety of methods appear to provide three-dimensional (3-D) reconstructions of blood vessels in a patient. Generally the results of such methods look very realistic. However, only a few produce a true 3-D reconstruction. We strongly suggest that for a true 3-D reconstruction of a blood vessel the following criteria should at least be fulfilled: (1) the arterial wall rather than the lumen must be reconstructed; (2) the spatially curved course of the vessel must be included; and (3) the orientation of local vessel wall characteristics, for example, plaque eccentricity, with respect to the luminal course must be correctly maintained. Currently, only methods combining biplane X-ray angiography and intravascular ultrasound imaging (IVUS) have succeeded in providing true 3-D reconstruction of a segment of a vessel. Accuracy of those reconstructions is derived from studies using phantoms having precisely known geometry. In patients, data on accuracy are more difficult to obtain. Nevertheless, a comparison can be made between the actual length of an IVUS pull-back trajectory and its reconstructed length showing relative differences of less than 3%. Further knowledge can be obtained by comparing simulated angiograms derived from the 3-D reconstruction with the real contrast angiograms. True 3-D reconstruction methods of the vessel wall and lumen, applicable in the individual patient, have become feasible and produce accurate results. Application of such a method will be helpful to understand immediate and long-term vessel remodelling induced by all types of catheter interventions and in the study of progression or regression of atherosclerotic wall disease.