Periodic practice is needed for newly trained robotic surgeons to maintain skills during periods of robotic inactivity. The current study was performed to determine whether virtual robotic skill maintenance can serve as an adequate substitute for practice on a surgical robot. Eleven surgical residents with no prior robotic training were trained to a level of robotic proficiency with inanimate models, including a needle driving pad, a running suture pad, and ring placement on a rocking peg board. After reaching proficiency, each resident was tested on a complex tissue closure task. For the next 8 weeks, the only robotic activity was biweekly virtual robotic skills maintenance. After 8 weeks, the residents performed the tissue closure task twice with the robot, followed by evaluation on the inanimate models used to reach proficiency. Repeated-measures statistical analyses were used to compare between the three tissue closure trials and between the final test at week 0 and the evaluation at week 8 for the other inanimate models. Time to complete the tissue closure task was more than 20 % lower for the second evaluation at 8 weeks than for the other two trials (p < 0.05). Residents maintained their skills for needle driving, but times for suture running and rocking peg board increased by more than 20 % at 8 weeks (p < 0.01). Virtual practice shows promise for maintaining robotic skills. Following a warm-up period, some skills may actually improve with biweekly virtual practice, but skill retention is selective, so further improvements are needed.