While solid‐state materials are commonly classified as covalent, ionic, or metallic, there are cases that defy these iconic bonding mechanisms. Phase‐change materials (PCMs) for data storage are a prominent example: they have been claimed to show “resonant bonding,” but a clear definition of this mechanism has been lacking. Here, it is shown that these solids are fundamentally different from resonant bonding in the π‐orbital systems of benzene and graphene, based on first‐principles data for vibrational, optical, and polarizability properties. It is shown that PCMs and related materials exhibit a unique mechanism between covalent and metallic bonding. It is suggested that these materials be called “incipient metals,” and their bonding nature “metavalent”. Data for a diverse set of 58 materials show that metavalent bonding is not just a superposition of covalent and metallic cases, but instead gives rise to a unique and anomalous set of physical properties. This allows the derivation of a characteristic fingerprint of metavalent bonding, composed of five individual components and firmly rooted in physical properties. These findings are expected to accelerate the discovery and design of functional materials with attractive properties and applications, including nonvolatile memories, thermoelectrics, photonics, and quantum materials. Metavalent bonding is introduced, defined in terms of physical properties, and identified in a class of inorganic solids. This includes phase‐change materials and thermoelectrics previously referred to as “resonantly” bonded. Metavalent bonding is shown to be a unique and fundamental bonding mechanism in solids: it is located between covalent and metallic interactions, but clearly distinct from both.