The GREATEST new T-SQL Function

If the title isn’t obvious enough, this blog post is about the new function GREATEST, and it’s sibling LEAST. But wait, how come there’s a new T-SQL function? There isn’t a new release of SQL Server! That’s right, those two new functions are released in Azure SQL DB, Azure SQL Managed Instance and Azure Synapse Analytics. In other words, cloud first, on-prem later. I learned about those functions when reading the blog post Your Top 10 Ideas for SQL Server by Brent Ozar, where he subtle tricks people in commenting about which features they want in SQL Server, and then he scolds them for not having submit a feature request 😉 Anyway, someone in the comments mentioned the discovery of the 2 new functions.

So what do they do exactly? You can look at GREATEST as a MAX function, but for columns instead of rows. Let’s demonstrate with an example. We got the following simple table:

SELECT 1 AS Col1,2 AS Col2,3 AS Col3
INTO #Test
UNION ALL
SELECT 6,4,5
UNION ALL
SELECT 10,-5,9;

SELECT *
FROM #Test;

To get the highest value out of the 3 columns for each row, we can now use the following statement:

SELECT GREATEST(Col1,Col2,Col3)
FROM #Test;

As you can see from the screenshot, Intellisense doesn’t recognize this great function (pun intended) yet. I really like this addition, as in ye olde times we had to write a lot of nested CASE/IIF functions to get the job done (or you could pivot/unpivot the data and use MAX if that’s your thing). And it only got more harder if more columns were involved. Also, it was one of the things that Oracle had, but SQL Server didn’t. So it gets easier to migrate PL-SQL to T-SQL 🙂

LEAST does the opposite of course:

The functions work with any comparable data type, such as dates:

And luckily, they ignore NULL values:

UPDATE: I’ve used “column” throughout this blog post, but “expressions” is more accurate. This can be a column, a subquery, a variable, a calculation and so on.