A self join is a database operation where a table is joined with itself, typically to compare rows within the same table. It is useful for querying hierarchical data, finding duplicates, or comparing rows based on certain conditions within the same dataset.