This is absolutely brilliant! I won't ask you - because you've been more than generous with your time already - but if I can:

- highlight any numbered headings

-autopopulate entries in the FindText array with 'any bold, non-highlighted strings of words in the document', ie definitions (but not headings since highlighted under the 1st step above)

- run your code

- generate a list of number of instances each term appears.

I will know in an instant which definitions are not used. My whole job revolves around definitions so this is amazing progress!