Don't see the reason to combine id, an identifier, and the 
[name] field as the primary key.  Even both ID and name are CKs, only one 
should be selected as the PK, while the AK could be enforced with a unique 
index.

In any case with composite primary keys, I would add an id identifier as 
the primary key of the table, then create a unique index on the combined 
PK.  Maybe not good for a modeling homework though. 

> Here is a short example that defines two fields as the primary keys of a 
table:
> 
> CREATE TABLE [dbo].[Table1] (
>          [ID] [int] IDENTITY (1, 1) NOT NULL ,
>          [Name] [char] (10) NOT NULL
> ) ON [PRIMARY]
> GO
> 
> ALTER TABLE [dbo].[Table1] WITH NOCHECK ADD
>          CONSTRAINT [PK_Table1] PRIMARY KEY  CLUSTERED
>          (
>                  [ID],
>                  [Name]
>          )  ON [PRIMARY]
> GO
> 
> 
> HtH
> 
> Imar
> 
> P.S. You could create the table in EM, than script your table to a SQL 
file 
> to see how the SQL should look.
> 
> At 05:34 PM 1/30/2001 +0000, you wrote:
> >I can create 2 primary key in a table with EM, but when I used SQL
> >language it doesn't work. I'll be grateful of any help.
>