Virtual reality (VR) applications are rapidly permeating fields such as medicine, rehabilitation, research, and military training. However, VR-induced effects on human performance remain poorly understood, particularly in relation to fine-grained motor control of the hand and fingers. We designed a novel virtual reality environment suitable for hand-finger interactions and examined the ability to use visual feedback manipulations in VR to affect online motor performance. Ten healthy subjects performed a simple finger flexion movement toward a kinesthetically-defined 45 target angle while receiving one of three types of VR-based visual feedback in real-time: veridical (in which the virtual hand motion corresponded to subjects' actual motion), or scaled-down / scaled-up feedback (in which virtual finger motion was scaled by 25% / 175% relative to actual motion). Scaled down-and scaled-up feedback led to significant online modifications (increases and decreases, respectively) in angular excursion, despite explicit instructions for subjects to maintain constant movements across conditions. The latency of these modifications was similar across conditions. These findings demonstrate that a VR-based platform may be a robust medium for presenting visuomotor discordances to engender a sense of ownership and drive sensorimotor adaptation for (retraining motor skills. This may prove to be particularly important for retraining motor skills in patients with neurologically-based movement disorders.