Before we jump into the core logic, let me explain you what are the use of this functions / property.
BTW, this is the table to be used in example:
FirstName varchar(10)
,MiddleName varchar(10)
,LastName varchar(10)
select ‘Ritesh’,‘A’,‘Shah’ UNION ALL
select ‘Roger’,NULL,‘Federer’ UNION ALL
select ‘Steffi’,NULL,‘Graf’
Select FirstName,LastName fromEMPTEST
ISNULL: This function replaces the NULL value with specified value given in function. Suppose we have probability to have Middle Name NULL in Employee table, we could do something like below given TSQL statement.
Select FirstName, ISNULL(Lastname,) as LastName From EMPTEST
If Lastname will be NULL, it will be replaced with blank in result set.
COALESCE: Basically COALESCE function will return first not null value from the given list, so we can use this function in place of ISNULL too, like this:
Select FirstName, COALESCE(Lastname,) as LastName From EMPTEST
NULL is nothing but the absent of value, it doesn’t even represent a blank or space or zero. When you try to concatenate two or more than two strings and any of the string is NULL, it will return the NULL only. See following example:
Select FirstName + ‘ ‘ + MiddleName + ‘ ‘ + LastName as FullName fromEMPTEST
You will get first record “Ritesh A Shah” as full name and remaining two records as NULL as those two records are having NULL value in its MiddleName field so concatenate results will be NULL only.
You can have solution for that with ISNULL & COALESCE.
Select FirstName + ‘ ‘ + ISNULL(MiddleName,) + ‘ ‘ + LastName asFullName from EMPTEST
Select FirstName + ‘ ‘ + COALESCE(MiddleName,) + ‘ ‘ + LastName asFullName from EMPTEST
In this solution, you have to wrap up all the fields all the fields with either ISNULL or with COALESCE if it has probability of containing NULL value. If you have many fields in one query which needs this wrapping, it is bit tedious for you as a developer. At this time, CONCAT_NULL_YIELDS_NULL property comes as a rescue.
CONCAT_NULL_YIELDS_NULL: this property controls concatenation of the string, like what should do if any of the string is NULL and it is being used in concatenation.
You can set this property at database level by ALTER DATABASE command or you can use this for your current session/connection.
Select FirstName + ‘ ‘ + MiddleName + ‘ ‘ + LastName as FullName fromEMPTEST
As I have already told you that above query returns two record with NULL value as FullName, you can do something like below TSQL to set CONCAT_NULL_YIELDS_NULL  for your batch.
Select FirstName + ‘ ‘ + MiddleName + ‘ ‘ + LastName as FullName fromEMPTEST
By default CONCAT_NULL_YIELDS_NULL  set with TRUE (ON) for every database but you can turn it OFF for your batch or for your database (by ALTER DATABASE command). It will save you from writing ISNULL and COALESCE command for many times in your query.
If you want to check whether your database is set to TRUE or FALSE for this property or your SESSION has this property TRUE or FALSE, you can use following TSQL.
SELECT DATABASEPROPERTYEX(‘YourDatabaseName’, ‘IsNullConcat’)
Personally I prefer to go for either ISNULL or COALESCE rather than CONCAT_NULL_YIELDS_NULL . There are few reasons for that.
–> CONCAT_NULL_YIELDS_NULL  is supported in even SQL Server 2008 R2 along with previous versions of SQL Server but as per Microsoft, They will remove this feature in new releases, it will be by default ON in new versions and you will not be able to set it OFF
 –> If you have this option in Stored procedure, it will force your Stored Procedure to recompile every time you execute your SP.
Apart from these, as long as possible, we should avoid tempering with default settingsof SQL Server unless you know what exactly your doing.
 Reference: Ritesh Shah
Note: Microsoft Books online is a default reference of all articles but examples and explanations prepared by Ritesh Shah, founder of
Ask me any SQL Server related question at my “ASK Profile
Microsoft SQL Server Blog. Fight the fear of SQL with Founder is Ritesh Shah


  1. Pingback: Useful functions in SQL Server

Comments are closed.