Now, that I am thinking of it more, I have this situation since whatever was created since a few months. If you recall correctly, there was a thread a few months ago and it ended up that I adjusted my ALTER TABLE ADD COLUMN syntax to include in one line the addition of a field and a replace of all actual records with a blank values. Before, I was adding the field and doing an UPDATE after. Now, that this is done in one line using that approach, it creates a constraint. I am not sure I like it. I think I will keep it as it was months ago by adding the field in one line following by another line to update all existing records with a blank value. This will avoid those constraints to be created. Because, I do not need them at the SQL Server level. This is all taken cared of automatically by the framework. We reached that situation because we wanted to initialize with a default values all the existing records. This applied of course for those tables that contain records at the time I execute the command to add a new column. If I use back that approach, I wouldn't even have to bother about checking for constraints and removing them, if any, when I would want to enlarge a column length for example.