Visuospatial abilities have been demonstrated to predict the performance of medical students in simulated endoscopy. However, little has been reported whether differences in visuospatial abilities influence the performance of senior endoscopists or whether their vast endoscopy experience reduces the importance of these abilities. Eleven senior endoscopists were included in our study. Before the simulated endoscopies in GI Mentor II (gastroscopy: case 3, module 1 and colonoscopy: case 3, module 1), the endoscopists performed three visuospatial tests: (1) pictorial surface orientation (PicSOr), (2) card rotation, and (3) cube comparison tests that monitor the ability of the tested person to re-create a three-dimensional image from a two-dimensional presentation as well as mentally manipulate that re-created image. The results of the visuospatial tests were correlated to the performance parameters of the virtual-reality endoscopy simulator. The percent of time spent with clear view in the simulated colonoscopy correlated well with the performance in the visuospatial PicSOr (r = -0.75, P = 0.01), card rotation (r = 0.75, P = 0.01), and cube comparison (r = 0.79, P = 0.004) tests. The endoscopists who performed better in the visuospatial tests also were better at maintaining visualization of the colon lumen. Those who performed better in the PicSOr test formed fewer loops during colonoscopy (r = 0.60, P = 0.05). In the technically less demanding simulated gastroscopy, there were no such correlations. The visuospatial tests performed better in endoscopists not playing computer games. Good visuospatial ability correlates significantly with the performance of experienced endoscopists in a technically demanding simulated colonoscopy, but not in the less demanding simulated gastroscopy.