As I said, _handleri really looks, as if it should be some
system variable. Maybe pointing to some prg or app. Then some
code like DO (_handleri) fails because of misconfiguration of
_handleri.

I thought I saw a definition of _handleri in the install
routine of amri. But that isn't so, there is something with
AMHIDE_RI, but not _handleri. Maybe that's a rest of your
previous RI code? I also didn't find someting about _handleri
with google r in the online (msdn) help of vfp9. I also
didn't find it in the form of the AmRI.pjx or in stdcode.prg
of AmRI.

Try making a copy of your database, remove any stored
procedures (concerning RI) and then start over with AmRI.

Does that help?

AmRI is very cautious with old code and may also not overwrite
all trigger calls. So if nothing is defined as insert trigger,
an old insert trigger call may still be done and lead to an
error.

So also deleting all trigger calls before creating the AmRI RI
code may help:
Delete Trigger On <tablename> for Delete|Insert|Update
for all tables of the database (ADBOBJECTS helps!).

That doesn't modify the stored procs themselves, it's only
manipulating what's written in the table tab in the table
designer as call for delete|insert|update trigger and that is
the call each of these triggers trigger.

If all else fails, maybe just put a SET STEP ON in the trigger
code for the table and see what's running wrong...

Today I got my copy of VFP9 and I'll try it myself, when I
have some time. This thread get's a little OT, so I'll contact
you via mail until next monday. Is that okay?

Bye, Olaf.