AIM: To investigate the relation between intrauterine pressures and volumes for virtual-reality-based surgical training in hysteroscopy. MATERIAL AND METHODS: Ten fresh extirpated uteri were insufflated by commercial hysteroscopy pump and imaged by computer tomography (CT) under intrauterine air pressure in distension-collapse cycles between 0 , 20 (150 mmHg), and 0 kPa, performing a CT scan at every step at about 2.7 kPa (20 mmHg). RESULTS: An initial threshold pressure to distend the cavity was avoided by introducing the insufflation tube up to the fundus. The filling and release phases of seven uteri that were completely distended showed the typical characteristics of a hysteresis curve which is expected from a viscoelastic, nonlinear, anisotropic soft tissue organ like the uterus. In three cases tightening the extirpated uterus especially at the lateral resection lines caused significant problems that inhibited registration of a complete distension-collapse cycle. Interpolated volumes for complete distended cavities and extrapolated for incomplete data sets, derived from the digitally reconstructed three-dimensional (3D) geometries, ranged from 0.6 to 11.4 mL at 20 kPa. These values highly correlate with the uterine volume (not insufflated) considering different biometric data of the uteri and patient data. Linear (R (2) = 0.66) and quadratic least-squares fits (R (2) = 0.74) were used to derive the formulas y = 0.069x and y = 0.00037x (2) + 0.036x, where x is the uterine volume in mL (not insufflated) and y is the cavity volume in mL at 20 kPa intrauterine pressure. CONCLUSIONS: Our experimental hysteroscopical setup enabled us to reconstruct the changes in volumes of insufflated uteri under highly realistic conditions in 3D. The relation between intrauterine pressure and cavity volume in distension-collapse cycles describes a typical hysteresis curve.