The USPTO has released a new version of the Manual of Patent Examination Procedure (MPEP) along with a reworking of the MPEP website to allow for better facilitate online access to the rules and regulations. The new site can be found at  The new style is also designed to allow the USPTO to quickly modify various sections so that the MPEP “can be updated within hours, rather than months.”  A new version of the TMEP is also available for trademark prosecution at

47 thoughts on “MPEP.USPTO.GOV

  1. 47

    Maybe they update 3.73(b) in the rules, since it still refers to patents (or update the Fed. Reg.). The whole AIA roll out has been rushed and is a complete embarrassment.

  2. 46

    Stratégie au rugby pour gagner du terrain

    Fly, that’s the meaning I’ll happily accept, better than the one about escaping from a hapless situation!

    “Botter” is really nice. Charmingly onomatopoeic, wouldn’t you say? Thanks for that.

  3. 45

    “… they’d write one, it would just be preposterous.”

    And that would differ from the typical examiner OA how?

  4. 44

    Fly, my contributions here are intended to shape opinion. If I announce there is a Golden Thread then perhaps, one day, there will be. Meanwhile, may I take it that you do approve of both “fair” protection for inventors and, for the public, legal certainty? If so, which is more important?

  5. 43

    “I borrowed that phrase from 6, who seems to think that jurors would be incapable of generating an OA after an entire trial”

    Nah, not incapable, they’d write one, it would just be preposterous.

  6. 41

    Ah mon cher Max, vous êtes un maître de l’art du botter en touche

    I can’t see the relationship between Art. 69 and my gentle poke. There was a passage in the old Guidelines (the EPO’s MPEP) suggesting what the examiner should say when the applicant invoked Art. 69 when there were outstanding objections under Art. 84. Haven’t found it yet the new edition. (Sheesh! 800 pages!)

    From the 2010 Case Law compendium, page 111:

    In T 223/05 the board held that the interpretation of the extent of the protection of a patent was not the task of the EPO, but, according to Arts. 64 and 69 EPC 1973, that of the national courts competent in procedures on infringement cases (see T 740/96, T 442/91). In particular, Art. 69 EPC 1973 did not offer any basis for reading into a claim features which could be found in the description when judging novelty (see T 1208/97).

  7. 40

    Thanks, but the person who wrote the thing didn’t bother to put the year down. I guess I could try the last 7 years or so. I’ll give it a whirl sometime. I thought about trying that already but didn’t think it worth the trouble. I figured if it only has 6 digits it had to be what comes after a year designation.

  8. 39

    Espacenet and Register are two very different animals.

    Espacenet provides worldwide patent information. It is a public interface to the EPOQUE search engine and EPODOC database, with access to the BNS document image repository for original document access. Records and BNS images are added continuously, improving coverage in time (the oldest document, US1, is about 185 years old) and in space (eg: UA, CN, SU, SG, EG, BR, MX, etc.) The data must be normalized to allow searching. Application numbers are thus generally of the format CCYYYYNNNNNNN, with CC designating the country or office, YYYY the year using four digits, and NNNNNN the serial number, whose exact composition will vary depending on the country. This causes a bit of a problem when you are looking for, say, a US application (SS/NNNNNN), which does not explicitly indicate the filing year. The underlying database does have the application serial in a non-normalized format, but it is clumsy to search. Examiners will rather use command scripts (“..doc us”), which will attempt to find all possible applications corresponding to a given serial number. A similar problem occurs with older Japanese publication numbers which are based on Emperor years (how often did I ask myself: “when did Hiro Hito die?”…)

    Register is an application exposing the data maintained in EPASYS, the EPO’s workhorse for managing the EPC procedure (PCT and national applications have systems of their own), and provides access to file images from Phoenix. It is normal IMO that the system is more accommodating that Espacenet. The internal examiner’s interface to Phoenix is even more tolerant, but much less slick in its presentation. It is not limited to published EP files, but also allows access to national applications and PCT files before publication.

    Compared with the EPO Register service, Public PAIR is horribly clunky.

    I understand that PAIR was acquired from the EPO and adapted. I find that it improved tremendously in reliability and availability since its introduction.

    It would be nice if PAIR did away with CAPTCHAs (at least from an EPO IP address), allowed deep links, and supported file monitoring.

    I often look at IDSes. Most of the stuff cited in these is pretty useless, or is a mere transcription of the EP Search Report references. (The EPO could offer to print Search Reports in IDS format for a fee…) Sometimes there is some NPL stuff cited by the applicant which sounds promising or very interesting (e.g.: industry reports, litigation evidence), but which cannot be supplied in PAIR. As an examiner I couldn’t see how I could ask the applicant to supply me with a copy of those documents, as it does not literally fall under the definition of Art. 124 EPC. Why couldn’t a privileged access be provided within the confine of Trilateral offices? EPO examiners have (an almost) one-click access to Japanese file wrappers, including an automated translation of the searcher’s internal notes to the JPO examiner. Surely something could be worked out, the enabling legal framework already exists.

  9. 38

    Many thanks Fly, especially for the snippet that Espacenet is pickier than Register. Outside the EPO interested circles, everybody (it seems to me) knows Espacenet but not Register. Me, I always get what I want, and quick, from Register.

    Compared with the EPO Register service, Public PAIR is horribly clunky.

  10. 37

    The solution? Easy. Here goes.

    The golden thread is the Art 69 Protocol. It’s imperative in Europe to give the inventor a “fair” scope of protection, constrained only by that other imperative: the public must be given “reasonable” legal certainty. So, EPO, keep doing what you’re already doing but with the highest respect for the imperatives of that Protocol.

    Now to the Americans. Remember “It’s good to talk”. That means dialogue. They don’t take kindly to lecturing. That means putting to the task people who do understand “where they are coming from”. Have your people talk in that way, with Team Rader, with Team Kappos, with the AIPLA and with the IPO.

    Point out that he who declines to show what his invention is can hardly complain if he doesn’t get for his invention a scope of protection in Europe that is “fair”.

    After that, naked self-interest will do the rest.

  11. 36

    Claims hopelessly lacking in novelty. So, WHAT’S THE INVENTION? I DUNNO (no Summary Section, no claims pitched at a sensible level of generality).

    It is extremely refreshing to see this admission from a representative.

    What’s your solution?

    Make the examiner figure out the invention for you?

    Fight a war of attrition?

    Send back the application to your correspondent?

  12. 35

    only 6 digits with an “N” at the end

    That doesn’t sound like anything I know. EP publication numbers passed the one million mark years before 2005, and “N” is not a valid document kind code in any country. (I checked in Espacenet, nothing comes up).

    It can’t be an EP application number either, there’s just not enough digits.

    If the application claims a foreign language priority, I suggest you have a look at the PRIO doc, the document could have suffered at the translator’s hand. (It can happen).

    The number could refer to a national publication number. A few smaller countries used sequential numbers in the 6 digit range, such as CH, AT, PL. (OK, Poland ain’t that small).

    I’d try searching by using the name of the applicant or of the inventors as a key.

  13. 34

    Monsieur Six,

    There is no such thing (yet) as an “EU” application!

    You either mean an EP (EPO) application, a WO (PCT), or a national one (FR, DE, NL, IT, etc.).

    EP application numbers (as marked on the wrapper) are of the format YYNNNNNN.X . YY designates the year (79..99,00..12) the file (notionally) arrived at the EPO. NNNNNN is a 6 digit application serial within the year, with the first one or two digits indicating the route it took. (Electronic filing at the EPO; paper filing dropped in the mailbox at The Hague, Berlin, or Munich; filing received by a national office and forwarded to the EPO; PCT application entering the regional phase, in the last case the year is the one of the international filing). X is a check digit determined using the Luhn algorithm. It can often be omitted.

    All applications are published 18 months after filing or the oldest claimed priority, whichever comes first, and if the required fees are paid. If your application number begins with 11 or 12, there’s more than a good chance it’s not available yet.

    Register (an implementation of Article 128 EPC) will accept application numbers in the format YYNNNNNN, YYNNNNNN.X, or EPYYYYNNNNNN. (I believe that EPYYNNNNNN will also work, as it did internally. Register is less picky than PAIR).

    Espacenet will only accept a format EPYYYYNNNNNN .

  14. 33

    One would hope that the PTO would continue to assign edition and revision numbers to the MPEP whenever it is updated revised.

  15. 32

    Interesting to note that when someone else is the target, MaxDrei is all shtts and grins.

    Another prisoner laid to rest.

  16. 31


    I believe that falls under the “Too Soft” (lacking credibility) category. It might be hard to explain in a report on the effect on Small Business to have your sources be Big Business.

  17. 30

    anon, they could ask the usual suspects, such as the leaders of the harmonization movement who for so long have pushed FTF? These usual suspects routinely state with utter conviction that the primary beneficiaries of the new act are the small inventor. So why even bother to ask the small inventor or startup their views? That would be a waste of time; and it might, Gd Forbid, rustle the feathers of the powers that be in congress who might have actually believed what the usual suspects told was the disinterested truth they were assured it was.

  18. 28

    What if you cite to an MPEP section in a Response and by the time the Examiner reads it, it’s gone or changed? Will the OED get after me for misrepresenting the MPEP? And how would I prove it?????

    If you’re citing from the online version, maybe it would be prudent to cite the date, including the time of day, on which you accessed the document.

  19. 27

    “EPO Examiners are required to search once and once only. They are required to search not only what the as filed claims recite but also what they think Applicant will cut back to, when faced with the EPO’s best art.”

    Sounds a lot like MPEP 904+. If you consider that an astute U.S. examiner has actually read those sections, the number of astute examiners at the USPTO is about 4.

  20. 26

    As you say, no need to answer. Still, I can’t help it.

    EPO Examiners are required to search once and once only. They are required to search not only what the as filed claims recite but also what they think Applicant will cut back to, when faced with the EPO’s best art. That’s why.

    Try amending down though, to something unexpected. That will provoke the responses:

    1. New matter. So, not admissible.

    and, even if no new matter:-

    2. Not searched. So, not examinable.

    Then you have to file a div (if you are still in time, of course). And even if you are in time, we have all heard just how grotesquely expensive that is reputed to be, at the EPO.

    Which is why savvy Applicants at the EPO flag up, at the outset, what technical feature combinations they want searched.

  21. 25

    Many thanks all gents, but I was unable to find anything looking like what I would have thought the ref was about. The number I was given is only 6 digits with an “N” at the end and it must be a newer ref because it is about a tech that would have come about in around 2005ish.

    In any event, I gave up after trying lots of permutations. I think it’s either a typo or something. In any event, nobody else around me is able to find it either so screw it.

  22. 24

    “It would not surprise me if USPTO Examiners spit out just about any old rubbish…”

    They’re taught, by their SPE’s and primaries, to do that regardless of the “quality” of the application. Issue a cr#ppy FAOM, issue an even cr#ppier FR, possibly issue an Advisory, and wait for RCE. If the applicant appeals, wave magic wand and determine that all of the previously unpersuasive arguments are now persuasive and re-open.

    Can you explain why EPO examiners, when examining “classic US-drafted form” applications still somehow manage to find D1 and stick with it throughout prosecution? Lots of EP applications claim priority to U.S. applications, and they are usually filed in the EPO “as is” without too much effort to account for EPO requirements (e.g. one can’t dicuss D1 in the background until the EP examiner identifies D1). Are EP examiners just that much more astute than U.S. examiners? (No need to answer, I already know.)

  23. 23

    Again I laughed. But here I will offer something of a defence of the difficulties USPTO Examiners face.

    The file I was on this morning began with a WO document in classic US-drafted form, that is to say:

    Field/Background/Brief Descrn of Drgs/Detailed Descrn/Claims

    Claims hopelessly lacking in novelty. So, WHAT’S THE INVENTION? I DUNNO (no Summary Section, no claims pitched at a sensible level of generality).

    It would not surprise me if USPTO Examiners spit out just about any old rubbish, as their FAOM, just to get back from the Applicant something on which they can get started. They might therefore be more astute than you suppose JJ.

    I should say that this is NOT how the rest of the world drafts patent applications. In the rest of the world, it really is NOT a good idea to refrain from telling the readers of the app as filed what the invention is.

  24. 22

    I borrowed that phrase from 6, who seems to think that jurors would be incapable of generating an OA after an entire trial, something which examiners are trained to do in about an hour: 1)text search with inexplicable Boolean logic; 2) cut and paste highlighted text from search results into OA generating software; 3) make sure all claims rejected; and 4) turn in for counting.

    I’ll let you know when I come across one of the more astute examiners at the USPTO. About as elusive as the snark.

  25. 21

    I laughed at that JJ. Thanks. I like the sound of that word “nubcake”. The “nub” of an invention is its core, its crux, so not very derogatory. In nubcake, where does “nub” come from and what does it mean? Is nubcake the new “muppet”?

    But I’m sure some of the more astute USPTO Examiners check out the EPO file. In my experience, soon after good art appears on the EPO file, the Examiner in China cites it, and even sometimes the one at the USPTO.

  26. 20

    “USPTO Examiners might find it helpful to check at the EPO…”

    Lulz. The PTO’s own super examiner, 6, the one who claims he can find knockout 102(b) prior art on any case with a five second Google search, can’t even figure out how to use the EP Register. Good luck having any of the other nubcakes checking it out.

  27. 18

    Mere speculation, but two of the more interesting reasons why the USPTO has missed the deadline for its report: Effects of First-to-File on Small Business – due to attempts to “sanitize” the draft report:

    Too Hard

    The draft is critical of First Inventor To File. This would not be surprising given that the main proponents have been the Large multinational firms, small firms are naturally far less interested in a World patent and that the traditional territorial US patent system is preferable.

    Too Soft

    The draft is so laudatory that it lacks any credibility.

    Goldilocks is busy.

  28. 17

    Oh good–now the PTO can change the MPEP in “hours.” Not that there’s anything left in the MPEP that helps argue in favor of patentability (after the past several years of deletion of such material), but in case anyone files a Response that does cite something useful they can get rid of it immediately. Shouldn’t there be some stability to the MPEP, so that one can at least count on the material being there from the time a Response is filed to the time the Examiner picks it up? What if you cite to an MPEP section in a Response and by the time the Examiner reads it, it’s gone or changed? Will the OED get after me for misrepresenting the MPEP? And how would I prove it?????

  29. 16

    link to

    This is where I start. Next step: input “EP” followed by a number (filing or publication) that is associated with the sought-for EPO filing.

    As AAA JJ helpfully points out, top left on the next screen are some useful tabs like “patent family”. The tab I use most is the one labelled “all documents”. That gives you all the docs on the EPO file, chronologically listed.

    And, of course, until the EPO lays open its file, all you can see in “all docs” is the international phase history of the foundation PCT filing, if there is one. Since the EPO allocates a filing serial number of its own series to every PCT filing, soon after the PCT is filed, there should be stuff to read under “all documents” as soon as you have a WO publication number to enter.

    USPTO Examiners might find it helpful to check at the EPO for i) the search report ii) the annex to the EPO search report, called the “opinion on patentability”.

  30. 15

    Ned Heller said:

    “I am waiting.

    What? Zero? Really?”

    With all due respect Mr. Heller, I am patiently waiting for you to perform “Integration Analysis” on Ultramercials claims, as requested three months ago.

    Perhaps now that “Integration Analysis” is officially in the MPEP, including sections and examples for Computer Programming, you will feel more comfortable in giving it a try?

  31. 14

    I find it amazing that we have a 9-0 Supreme Court decision relying on “Integration”. Followed by USPTO Official Guidelines on Integration Analysis. And now the addition of Integration Analysis in the MPEP itself including sections on computer programming, just like I proposed! Yet the blogs most vocal commenters all refuse to discuss the subject and have even run away from it. Some going to the extreme of lying and saying they talked about “Integration Analysis years ago, just to avoid having to deal with the subject! Yet there it is in black and white. It does appear that certain segments of the community fear “integration” like the South feared the end of segregation. But just like Civil rights, the laws and the times are changing. Read and learn my friends, read and learn.

    2100> 2106> 2106.01 – Subject Matter Eligibility Analysis of Process Claims Involving Laws of Nature (e8r9)

    …” 101 USPQ2d at 1966. To show integration, the additional elements or steps must relate to the natural principle in a significant… Along with integration, the additional steps must be sufficient to ensure that the claim amounts to significantly… …e.g., data gathering or data storage, would not show integration. For example, a machine that is simply incidental to execution of…”

    link to

  32. 13

    In 50-51, a draft statute was distributed to every patent law organization for comment. A coordinating committee was established. Many oraganization provided comments. A redrafted bill was distributed to every registered patent attorney for comment. It was distributed to every member of the APLA, who voted on the new bill, section by section, with comments.

    Testimony of Harry Ashton, link to

    (Of course, industrial organizations separately lobbied the bill as well.)

    Now contrast that what happened with the AIA. Were our views as patent attorneys sought out? Did we have a right to vote?

    The AIPLA was an active lobbyist on behalf of the bill. Even so, it did not, as far as I know, seek the advice of its own 500+ member patent law committee, nor did it seek acvice or approval from its members on any feature.

    If, in the end, it all works out, bully for that. But what happened here was downright undemocratic in the extreme. It leaves a very bad taste, and I can only hope that the AIA is held unconstitutional because of some of its more controversial provisions, such as eliminating 102(f) so that a valid patent could issue to a non inventor.

  33. 12

    When you enter a search using the application number, it should be the letters EP followed by either 8 or 10 digits. If 8 digits, the first 2 digits are the year of filing (e.g, 09, 10, 11, etc.) and the last 6 are for the application number (add preceding zeroes if you don’t have enough). If using 10 digits, the first 4 are for the year (e.g., 2009, 2010, 2011). Either way should work, and there is no reason why you should not be able to find a published EP application if you have the correct application number.

  34. 11

    Perhaps it hasn’t asked the Canadians who have reported adverse affects.

    Now, lets count the venture capitalist firms in Germany.


    I am waiting.

    What? Zero? Really?

  35. 10

    OT, but of note:

    The USPTO has missed the deadline for its report: Effects of First-to-File on Small Business.

  36. 8

    If you go to espacenet and do a number search of your published EP application (i.e. the publication number, not the application number), under the “Bibliographic Data” heading you will see the “EP Register” link. Click on that. A new window will open. On the left you can click on the “All Documents” tab.

    Like PAIR, you can only view published applications. As there is no non-publication option at the EPO, any app filed claiming priority more than 18 months ago will be published.

  37. 7

    Thank you sir I did already try espacenet. It doesn’t return anything for my number, I fear that it may be unpublished. But I’m also not convinced that espacenet has the ability to find all applications by app number.

  38. 5

    Maxdrei, or anyone who can assist. Just how do you find an EU patent application that you only know the application number of and don’t even know which office it was filed in? Is there a way?

    We have PAIR in the US for this, but what do you guys have?

  39. 3

    “MTEP” is a rather funny typo on a number of levels…as in “Empty”, etc. In fact, one wonders if that is why they decided to call it the Trademark Manual of Examining Procedure in the first place. But Dennis you probably ought to correct it to read “TMEP” lest you offend the trademark practitioners. 😉

  40. 2

    So how come it still has the BPAI and not the PTAB? I guess that’s could be updated in hours, not will be.

  41. 1

    can be updated within hours, rather than months

    Only half joking, but is this (additionally) so if the AIA has an injunction placed against it for constitutional grounds, certain sections can be more easily segragated?

Comments are closed.