RATIONALE AND OBJECTIVES: The aim of this study was to assess if graphical user interface screen-capture software applied to a magnetic resonance (MR) hardware console could nonintrusively allow the analysis of discrepancies between expected and experienced MR table time. MATERIALS AND METHODS: Individual MR examination acquisition processes were captured in audio-video interleave streams for 10 patients who underwent comprehensive liver MR imaging. These audio-video streams were dissected into periods of true image data acquisition, scanner activity without direct image data acquisition, and scanner inactivity. RESULTS: Total expected time required for standardized liver MR image acquisition was estimated at 15 minutes. Experienced table times varied highly, ranging from 19:00 to 58:08 minutes. Image data acquisition occurred during approximately 58% (range, 43.3%-73.7%) of overall table time. Image data were obtained approximately 77% (range, 65.6%-87.0%) of the time the scanner spent active. CONCLUSION: Graphical user interface screen-capture software installed on an MR console nonintrusively revealed significant periods of table time spent not obtaining true image data and explained discrepancies between expected and experienced MR table times. Table-time calculations using Digital Imaging and Communications in Medicine image headers and scanner-logged time stamps are underestimations of true table time because they do not take into account some scanner activities not directly leading to image formation.