In Windows 2000, we generate pdfs from Frame 6.0 files, then use the Acrobat 5.0 Catalog feature to create a searchable index for 200+ documents. This usually works great. Recently, Search lost the ability to return hits for terms that include hyphens or underscores, but only those instances from one particular doc, which was recently reformatted in Frame in a new template. This same template is applied to other Frame files whose pdf output is indexed and searched successfully for the very same terms. We can't determine what 'shut off' access to these special character terms in this one doc -- is it a Frame thing, or an Acrobat thing?