Not sure what you mean by "check the whole database"?  I can't see your data
structure, so I don't have a sense of how you've organized the data, nor if
you've used any indexing.

If you want Access to check for duplicates, the spelling of all fields would
have to be an EXACT match.  Unless, of course, you are willing to specify in
code to Access how to determine "close match".  In the example you offer,
Access would not, without more "training", be able to spot Smith @ 1 Avenue
Rd as a duplicate.

Is there a chance you could USB (use someone's brain)?  For example, if your
system could look up name/address information, given a minimum of data (say,
the first three letters of the surname), would that allow your users to
visually determine that a customer already existed (and select that
customer)?

As an aside, when your customer Smith moves to 2 Avenue Road, or across
town, or across the country, how would Access or you or your users know it
was the same/duplicate customer?

There's a chance your goal of eliminating duplicates is too costly to
achieve.  Perhaps you will settle on reducing the percentage of
duplications...

Good Luck!

Jeff Boyce
<Access MVP