fhaist- We had another person here having the same pdf problems as I had been (and that you seem to be having). I suggested that she download the update before I came to help her and that seemed to have made the problem dissapear. The update states that "PDF imports now search PubMed to complete citation data with additional fields not found in CrossRef. " So, it’s possible that this was the key for us. I hope they have an update for you soon.
I want to update this post because all issues appear to be resolved. Thanks to the very helpful efforts of Jason Rollins at Thomson Reuters. It seems that the source of the problem for X4 on the Mac was that the default CrossRef (EBSCO) connection file did not connect to a reliable database. Dr. Rollins supplied a new CrossRef connection file and now the PDF import works like a charm. I can simply drop a PDF onto the EndNote icon on my dock and the ref is imported automagically. I don’t think Dr. Rollins would mind, so I am posting the CrossRef file that he sent me. This just needs to be placed in your Connections directory. He also mentioned that the X4.0.2 update uses the PubMed database in addition to CrossRef, hence that update can improve the import function for those references in PubMed. The update should be available for the Mac soon.
Again, many thanks to Dr. Rollins on his very quick and helpful support and to make this feature work with the EndNote on Mac.
CrossRef.enz (4.68 KB)
slpscientist - Thanks for the post. I saw that just after I posted my update. I’m not sure if the new CrossRef.enz file is included in the update. If not, everyone should download that and install it so the CrossRef part of the DOI searches will work.
FWIW, I’ve found that some venders are still not posting Searchable PDF’s to their databases, especially for older articles. Even if the imaged based (non searchable) PDF contains a DOI in text, endnote can’t find it. The same would be true for scanned PDF’s. We’ve taken to running all our image-based PDF’s through Adobe’s OCR feature in the full version of Acrobat. The purpose is 3-fold. First, if you use their “ClearType” feature it shrinks the PDF (sometimes up to 50%). Second, it renders them searchable (it’s OCR, so it’s never 100% accuracy, but it’s good enough) which allows us to use the Search Field in Endnote. Third, if there is DOI information in our document, endnote can now (sometimes, see note on OCR above) pick it up.
Why “Digital Object Identifier” wouldn’t work? Endnote should really search for both “DOI” and “Digital Object Identifier”. All of my IEEE papers don’t have the string “DOI” 
Any plan for this update?
I just upgraded to X4.02 for Windows 7. Importing directly from the PDF is one of the features that I wished for all through my PhD!
I tried it on doi: 10.1152/physiol.00049.2010
I used the import option from the toolbar in Endnote, and filtered as PDF. It imported perfectly! I know that’s only 1 from 1, but I’m inpressed. Now to try the Mac version on my superviser’s MBP…
Just one question: With all of these papers attached to the library now, will the software start using up all of my RAM just to run?
Once the PDFs are indexed (which can take a bit of time, if you are upgrading to X4 with a library with lots of PDFs, it shouldn’t degrade performance, as they are pointers to the files in subfolders and accessed only when triggered. (in my experience with X4.0.2 and Word 2003 on Windows XP).