Advances in virtual reality technologies have made digital social interaction increasingly common, yet it remains unclear whether virtual humans – or avatars - can ever attain the status of real people. While it might be assumed that improving graphical realism will narrow the gap between virtual and physical social agents, the recent discovery of the Medusa effect suggest that the inherent abstraction in presenting humans as avatars creates a reality boundary that may be difficult to overcome. In my PhD, I will examine whether this phenomenon extends to virtual reality to understand the psychological processes that shape when and how people experience avatars as ‘real’. Across three research stages, I wish to examine how representational level, immersion, and emotional responsiveness shape the realness perception of avatars. First, I will establish whether the Medusa effect occurs in virtual reality by comparing judgements of a live person, their avatar, and a digital image of that avatar presented within the virtual scene. I will then examine whether prolonged immersion in virtual reality alters participants’ sense of what feels ‘real’, testing whether they continue to rely on the physical world as their anchor or begin to treat the virtual environment as the primary level of reality. Finally, I will assess whether emotionally responsive avatars reduce abstraction costs and enhance perceived agency, experience, and realness. This will provide a theoretical account of mind perception in humans and avatars, and inform the development of virtual systems capable of supporting human-like social interaction.