Best way to obtain records that are NOT in another table

Question

I have two rather large tables in oracle. An Account table that has millions of rows. Each account may be enrolled into a particular program and therefore can also be in an Enrollment table, also with millions of rows. I'm trying to find the most optimal way to find any accounts in ACCOUNT that are NOT in the Enrollment table.
I was doing something like this:
select /*+ index(ACCOUNT idx_acct_no) */
a.acct_no
from ACCOUNT a
where a.acct_no not in (Select e.acct_no from ENROLLMENT e);
This takes a VERY long time to execute, even though I am using the index.
I even tried to use the PK on the ACCOUNT table, as it is also a FK on the ENROLLMENT table as such:
select a.acct_no
from ACCOUNT a
where a.id not in (Select e.id from ENROLLMENT e);
this too takes too long to get back (if at all).
Is there a better way to do this selection please?

Answer 1

Well if you have the energy to type in the whole list, the syntax you've given will work, unless you blow the permitted number of elements.
But a practical solution would be to turn the list into a table. You still haven't got the hang of this "giving us enough information" concept, so let's presume:
(1) you're on a version of the databasse whoch is 9i or higher
(2) you have this list in a file of some sort.
In which case use an external table or perhaps a pipelined function to generate output which can be used in a SQL statement.
If neither of these solutions works for you please provide sufficient information for us to answer your question correctly. Your future co-operation is appreciated.
cheers, APC