Number 2 is faster. For number 1, the database needs to query for the
columns. For number 2, the columns are already specified.

If CustID is not nullable, then you'll get the same results.

However, I think you'll get different behavior if you have a nullable
column.. and you specify it.

Test it to make sure.

declare @t table (EmpID int not null  , DOB datetime null )
set nocount on
insert into @t (EmpID , DOB) values ( 123 , GETDATE() )
insert into @t (EmpID) values (456 )
set nocount off

SELECT   COUNT ( *)  FROM  @t
select count(dob) from @t

Just throwing it out there  .......  as a small gotcha.

Hello krishjaya,

You can use SQL Clients such as Management Studio, EMS or smth else to see
Exec plan of your query.
A lot of SQL clients can analyzy sql query and show you the execution plan
of your queries and time on each state of execution.

k> I have to use large volume SQL DB.
k>
k> My Table name is 'Customer' and has 52 fields with heavy data.
k>
k> Which SQL command is more fast? & why?
k>
k> 1.   SELECT   COUNT (*)  FROM Customer   WHERE CustId =
k> '1212322321313';
k> 2.   SELECT   COUNT ( CustId )  FROM  Customer  WHERE  CustId =
k> '1212322321313';
k> Thanx
k>
---
WBR,
Michael  Nemtsev :: blog: http://spaces.msn.com/laflour

"At times one remains faithful to a cause only because its opponents do not
cease to be insipid." (c) Friedrich Nietzsche

Ok Thanx for comments

Then can u explain me what is deference between

SELECT   COUNT ( * )  FROM Customer
SELECT   COUNT ( 0 )  FROM Customer

Thnx

KJ