This is probably a PHP misconfiguration that's causing PDFs to be incorrectly identified. OJS uses a few methods to determine file types: first PHP's mime_content_type function, and second the Unix standard "file" tool (if mime_content_type isn't available). Search this forum for mime_content_type for some suggestions; you may have to write a quick test script to determine why PDFs are being incorrectly identifed.
Open Journal Systems Team
Don't miss the First International PKP Scholarly Publishing Conference
July 11 - 13, 2007, Vancouver, BC, Canada