Background The influence of fatigue on residents' performance in laparoscopy was prospectively assessed through a computer-based virtual reality simulation (VRS) model. Study Design Twenty-six residents (14 novices, 12 experienced) were recruited. In the first stage, each participant was initially tested on 8 VRS-based tasks. In the second, run-in stage, each resident had 8 hours of hands-on practice of the specific tasks chosen. Finally, participants were evaluated before and after 24 hours on call. For each task, a set of parameters reflecting proficiency, efficacy, and safety were documented. Results In most of the tasks assessed, sleep deprivation had a significant deleterious effect on the performance of residents, both in terms of efficiency (time to complete the task), and safety (errors). These observations were more pronounced among novices. For example, in camera manipulation at a 30-degree angle, the total time to complete the task was slower after sleep deprivation (novices: sleep deprivation = 136 seconds, control = 119 seconds; experienced: sleep deprivation = 112 seconds, control = 103 seconds; p = 0.03). Moreover, accuracy rates were lower after sleep deprivation: in the "grasping and clipping" task, a lower accuracy rate after sleep deprivation was noted (novices: sleep deprivation = 82.8%, control = 87.9%; experienced: sleep deprivation = 87.6%, control = 90.8%; p = 0.05). Conclusions In this prospectively controlled study we observed reduced efficiency and safety in the performance of residents after sleep deprivation. Using an innovative study design, we eliminated the learning curve bias. Compared with novices, experienced residents performed relatively better after a night shift, and this may be attributed to better adaptation to sleep deprivation.