This study explored virtual-reality (VR) monitoring, the decision process by which people discriminate memories of real and virtual events. In a study phase, subjects touched 10 real objects with their real finger and touched 10 virtual objects (visual only, no tactile feedback) with their cyberfinger in VR. One week later they took a real versus virtual versus new source identification test. After the source identification test, subjects rated phenomenal qualities associated with each memory, using a Virtual-Real Memory Characteristics Questionnaire (VRMCQ). For old items, results from the VRMCQ are consistent with the idea that VR monitoring draws on differences in qualitative characteristics of memories for perceived and virtual events/objects (consistent with Johnson and Raye). However, subjects also reported similar qualities associated with their false memories for new items, suggesting that they sometimes infer/reconstruct the qualities a memory should have, based on their decision of its source of origin. Furthermore, VR monitoring might prove useful as a sort of Turing test of how convincing the virtual world is, and the VRMCQ can identify which qualities of the virtual experience (e.g., color) require improvement. Examples of applications are discussed.