A pre-existing belief about something. In Bayesian Statistics, the prior refers to the Distribution of potential values that one believes a system may hold. After observing data, a prior belief can be updated to a Posterior belief.