Populating a Table

Hello,

I'm sure I'll be flagged as a moron for asking this question, so be it.

I have some tables of info with no primary keys.  As a result the field I want to key on is repeated.  My plan is to do
aDISTINCT select from the un-keyed table into another table that does have constraints. 

Here is what I thought would work:

INSERT INTO table2
 SELECT DISTINCT(key), name, field1, field2
 FROM table1;

When I issue this statement it tells me that you cannot put in duplicate values into my key field.  Why does it do
that? If I issue the following statement I get exactly what I want: 

SELECT DISTINCT(phone) from table1;

Am I using the DISTINCT function incorrectly?  Is it taking all of the field and using them to create a distinct row?

Any help would be appreciated.

Chris.