Glass is everywhere. We use and are surrounded by glass objects which we then naturally consider as a distinct state of matter. Yet, glass as we know it is obtained by rapidly cooling a liquid below its melting point to avoid crystallisation. The viscosity of this supercooled liquid increases by many orders of magnitude upon cooling, until the liquid becomes arrested on experimental timescales below the glass “transition” temperature. From a structural viewpoint, the resulting glass very much resembles the disordered liquid, but from a mechanical viewpoint, it is as rigid as an ordered crystal. Does glass qualify as a separate state of matter? I will provide a pedagogical perspective on this question using basic statistical mechanical concepts. I will review recent theoretical results suggesting why and how an “ideal” glass can be defined as a separate equilibrium state of matter and discuss modern computer simulations trying to analyse this glass state. I will close with some experimental perspectives.