The NULL keyword specifies the null value. When assigned to a column of any data type, the NULL keyword forces the column to be set to null because the NULL keyword has no data type. For example: SQL> -- List all employees in the database and all potential employees SQL> -- SQL> SELECT employee_id, last_name, first_name cont> FROM employees cont> UNION cont> SELECT NULL, last_name, first_name cont> FROM candidates; . . . 00418 Blount Peter 00435 MacDonald Johanna 00471 Herbener James NULL Boswick Fred NULL Schwartz Trixie NULL Wilson Oscar 103 rows selected The NULL keyword is distinct from the IS NULL predicate which tests for null values of an expression. When testing an expression for null, using equality with NULL (for example, a predicate when testing an expression as shown in the following example: SQL> SELECT e.last_name, j.job_end cont> FROM employees e, job_history j cont> WHERE e.employee_id = j.employee_id cont> AND j.job_end IS NULL; E.LAST_NAME J.JOB_END Smith NULL O'Sullivan NULL Hastings NULL . . . In some cases, the NULL keyword may have a data type of CHAR(31) when used in a query that requires a data type; such as in arithmetic and character expressions. In such queries, the assumption of the CHAR data type may cause an incompatibility error. If this occurs, use the CAST function (for example, CAST (NULL AS data-type)) to change NULL to a compatible data type for the query.