Originally posted by jmodic
The main difference is if you need to reference the unique constrainted column with some foreign key. For that purpose you need unique CONSTRAINT, unique INDEX will not be enough.

So for ensuring uniqueness both methods will do the same job. but if you need a foreign key referencing that unique column (or combination of columns) you need UNIQUE KEY constraint.

Doesnt a foreign key reference need a primary key to point at, and not a unique constraint. (even though primary key is a unique constraint also).

Also will creating a unique constraint also automatically create a index.

Thanjs