Time is pervasive of the human way of approaching reality, so that it has been widely studied in many research areas, including Artificial Intelligence (AI) and relational Temporal Databases (TDB). Indeed, while thousands of TDB papers have been devoted to the treatment of determinate time, only few approaches have faced temporal indeterminacy (i.e., “don’t know exactly when” indeterminacy). In this paper, we propose a new AI-based methodology to approach temporal indeterminacy in relational DBs. We show that typical AI techniques, such as studying the semantics of the representation formalism, and adopting symbolic manipulation techniques based on such a semantics, are very important in the treatment of indeterminate time in relational databases.