This documentation is archived and is not being maintained.
Using Variables and Parameters (Database Engine)
Transact-SQL has several ways to pass data between Transact-SQL statements. These include the following:
Transact-SQL local variables.
A Transact-SQL variable is an object in Transact-SQL batches and scripts that can hold a data value. After the variable has been declared, or defined, one statement in a batch can set the variable to a value and a later statement in the batch can get the value from the variable. For example:
USE AdventureWorks; GO DECLARE @EmpIDVar int; SET @EmpIDVar = 1234; SELECT * FROM HumanRresources.Employee WHERE EmployeeID = @EmpIDVar;
Note: The maximum number of local variables that can be declared in a batch is 10,000.
A parameter is an object used to pass data between a stored procedure and the batch or script that executes the stored procedure. Parameters can be either input or output parameters. For example:
USE AdventureWorks; GO CREATE PROCEDURE ParmSample @EmpIDParm int AS SELECT EmployeeID, Title FROM HumanResources.Employee WHERE EmployeeID = @EmpIDParm GO EXEC ParmSample @EmpIDParm = 109 GO
Applications use application variables and parameter markers to work with the data from Transact-SQL statements.
The application programming languages such as C, C++, Basic, and Java have their own variables for holding data. Applications using the database APIs must move the data returned by Transact-SQL statements into application variables before they can work with the data. This is typically done using a process called binding. The application uses an API function to bind the result set column to a program variable. When a row is fetched the API provider or driver moves the data from the column to the bound program variable.
Parameter markers are supported by the ADO, OLE DB, and ODBC-based database APIs. A parameter marker is a question mark (?) placed in the location of an input expression in a Transact-SQL statement. The parameter marker is then bound to an application variable. This allows data from application variables to be used as input in Transact-SQL statements. Parameter markers also let stored procedure output parameters and return codes be bound to application variables. The output data is then returned to the bound variables when the procedure is executed. The DB-Library API also supports binding stored procedure parameter and return codes to program variables.