Relational databases consist of many objects, and the most common of these is a table. A table is a logical structure (relation) for storing and retrieving rows of data. It is defined by its columns which are represented by the data types.

Every table should have a primary key as a unique representation of a single row. It is usually a single column that either hosts a sequential number or a unique identifier that would never be used more than once. Tables can also contain two or more columns that together represent the unique row of the table. A table can also host foreign keys which describe its table reference, or relation to another table. For example, a Students table has a unique StudentID column, and the same StudentID is then represented in a Grades table to tie the actual grade to the student. This allows for one student to have none, one, or many grades, and requires that student record to exist before the grades can be entered.