Long count(*) time

Hello -

I'm running PG 8.2 on an 8-processor 16G Unix machine.  The machine is dedicated to the db, and only 5 threads/processors are busy.  The following query takes 70 seconds to execute. 
     select count(*) from url_list_url;
There are 64,219,173 rows in the table.  The table consists of an integer field and a text field.  The average length of the text field is 50 characters.  There are btree indexes on both fields.  The integer field is the key.

70 seconds seems to be a long time for this kind of query.  Is this normal?

Thanks.

David