Logical functions - GREATEST (Transact-SQL)
Applies to: 
SQL Server 2022 (16.x)
Azure SQL Database
Azure SQL Managed Instance
Azure Synapse Analytics
SQL analytics endpoint in Microsoft Fabric
Warehouse in Microsoft Fabric
This function returns the maximum value from a list of one or more expressions.
Transact-SQL syntax conventions
Syntax
GREATEST ( expression1 [ , ...expressionN ] )
Arguments
expression1, expressionN
A list of comma-separated expressions of any comparable data type. The GREATEST function requires at least one argument and supports no more than 254 arguments.
Each expression can be a constant, variable, column name or function, and any combination of arithmetic, bitwise, and string operators. Aggregate functions and scalar subqueries are permitted.
Return types
Returns the data type with the highest precedence from the set of types passed to the function. For more information, see Data Type Precedence (Transact-SQL).
If all arguments have the same data type and the type is supported for comparison, GREATEST returns that type.
Otherwise, the function will implicitly convert all arguments to the data type of the highest precedence before comparison and use this type as the return type.
For numeric types, the scale of the return type will be the same as the highest precedence argument, or the largest scale if more than one argument is of the highest precedence data type.
Remarks
All expressions in the list of arguments must be of a data type that is comparable and that can be implicitly converted to the data type of the argument with the highest precedence.
Implicit conversion of all arguments to the highest precedence data type takes place before comparison.
If implicit type conversion between the arguments isn't supported, the function will fail and return an error.
For more information on implicit and explicit conversion, see Data Type Conversion (Database Engine).
If one or more arguments aren't NULL, then NULL arguments are ignored during comparison. If all arguments are NULL, then GREATEST returns NULL.
Comparison of character arguments follows the rules of Collation Precedence (Transact-SQL).
The following types are not supported for comparison in GREATEST: varchar(max), varbinary(max) or nvarchar(max) exceeding 8,000 bytes, cursor, geometry, geography, image, non-byte-ordered user-defined types, ntext, table, text, and xml.
The varchar(max), varbinary(max), and nvarchar(max) data types are supported for arguments that are 8,000 bytes or less, and will be implicitly converted to varchar(n), varbinary(n), and nvarchar(n), respectively, prior to comparison.
For example, varchar(max) can support up to 8,000 characters if using a single-byte encoding character set, and nvarchar(max) can support up to 4,000 byte-pairs (assuming UTF-16 character encoding).
Examples
A. Return maximum value from a list of constants
The following example returns the maximum value from the list of constants that is provided.
The scale of the return type is determined by the scale of the argument with the highest precedence data type.
SELECT GREATEST('6.62', 3.1415, N'7') AS GreatestVal;
GO
Here's the result set.
GreatestVal
--------
7.0000
(1 rows affected)
B. Return maximum value from a list of character constants
The following example returns the maximum value from the list of character constants that is provided.
SELECT GREATEST('Glacier', N'Joshua Tree', 'Mount Rainier') AS GreatestString;
GO
Here's the result set.
GreatestString
-------------
Mount Rainier
(1 rows affected)
C. Return maximum value from a list of column arguments
This example returns the maximum value from a list of column arguments and ignores NULL values during comparison. This sample uses the AdventureWorksLT database, which can be quickly installed as the sample database for a new Azure SQL Database. For more information, see AdventureWorks sample databases.
SELECT P.Name,
P.SellStartDate,
P.DiscontinuedDate,
PM.ModifiedDate AS ModelModifiedDate,
GREATEST(P.SellStartDate, P.DiscontinuedDate, PM.ModifiedDate) AS LatestDate
FROM SalesLT.Product AS P
INNER JOIN SalesLT.ProductModel AS PM
ON P.ProductModelID = PM.ProductModelID
WHERE GREATEST(P.SellStartDate, P.DiscontinuedDate, PM.ModifiedDate) >= '2007-01-01'
AND P.SellStartDate >= '2007-01-01'
AND P.Name LIKE 'Touring %'
ORDER BY P.Name;
Here's the result set. LatestDate chooses the greatest date value of the three values, ignoring NULL.
Name SellStartDate DiscontinuedDate ModelModifiedDate LatestDate
-------------------- ----------------------- ------------------- ----------------------- -----------------------
Touring Pedal 2007-07-01 00:00:00.000 NULL 2009-05-16 16:34:29.027 2009-05-16 16:34:29.027
Touring Tire 2007-07-01 00:00:00.000 NULL 2007-06-01 00:00:00.000 2007-07-01 00:00:00.000
Touring Tire Tube 2007-07-01 00:00:00.000 NULL 2007-06-01 00:00:00.000 2007-07-01 00:00:00.000
(3 rows affected)
D. Use GREATEST with local variables
This example uses GREATEST to determine the maximum value of a list of local variables within the predicate of a WHERE clause.
CREATE TABLE dbo.Studies (
VarX VARCHAR(10) NOT NULL,
Correlation DECIMAL(4, 3) NULL
);
INSERT INTO dbo.Studies
VALUES ('Var1', 0.2),
('Var2', 0.825),
('Var3', 0.61);
GO
DECLARE @PredictionA DECIMAL(2, 1) = 0.7;
DECLARE @PredictionB DECIMAL(3, 1) = 0.65;
SELECT VarX,
Correlation
FROM dbo.Studies
WHERE Correlation > GREATEST(@PredictionA, @PredictionB);
GO
Here's the result set. Only values greater than 0.7 are displayed.
VarX Correlation
---------- -----------
Var2 .825
(1 rows affected)
E. Use GREATEST with columns, constants, and variables
This example uses GREATEST to determine the maximum value of a mixed list that includes columns, constants, and variables.
CREATE TABLE dbo.Studies (
VarX VARCHAR(10) NOT NULL,
Correlation DECIMAL(4, 3) NULL
);
INSERT INTO dbo.Studies
VALUES ('Var1', 0.2),
('Var2', 0.825),
('Var3', 0.61);
GO
DECLARE @VarX DECIMAL(4, 3) = 0.59;
SELECT VarX,
Correlation,
GREATEST(Correlation, 0, @VarX) AS GreatestVar
FROM dbo.Studies;
GO
Here's the result set.
VarX Correlation GreatestVar
---------- --------------------- ---------------------
Var1 0.200 0.590
Var2 0.825 0.825
Var3 0.610 0.610
(3 rows affected)