I argue for a novel answer to the question “What is hope?”. On my view, rather than aiming for a compound account, i.e. analysing hope in terms of desire and belief, we should understand hope as an irreducible concept. After criticizing influential compound accounts of hope, I discuss Segal and Textor's alternative of describing hope as a primitive mental state. While Segal and Textor argue that available developments of the standard definition do not offer sufficient conditions for hope, I question (...) the deep‐seated idea that desire and belief are even necessary conditions for hope. My suggestion is that we should take seriously the fact that we hope in a great variety of ways and should question the search for elements that are common to all cases. A promising alternative follows the Wittgensteinian idea that cases of hope are related in terms of family resemblance, i.e. are multiply realizable on the ontological level while falling under the non‐definable concept of hope. (shrink)