Challenging USPTO Decisions in the Courts

Last year the Board of Patent Appeals and Interferences decided over 7,000 ex parte cases. In about half of those decisions, the examiner rejections were entirely affirmed. After losing at the Board, an applicant has two primary avenues for challenging the Board’s decision: (1) Appeal to the Court of Appeals for the Federal Circuit under 35 U.S.C. § 141; or (2) File a civil action in district court under 35 U.S.C. § 145.

By the time that the BPAI appeal is complete, the average applicant will have spent over $25,000 on the application and prosecution process. That substantial figure is an indicator that the potential patent right is probably seen as valuable. It turns out, however, that that well under 1% of applicants who have a right to further action actually exert their right to appeal or to a civil action.

There are several likely reasons why the rate is so low:

  • BPAI Rehearing: The first reason for low appeals is likely that the BPAI has been fairly good at granting rehearing requests when there is a clear and easily identifiable mistake in their opinion.
  • Low Likelihood of Success: Under Zurko, the Federal Circuit is required to give deference to the USPTO’s findings of facts and only overturn those findings when “unsupported by substantial evidence.” This standard is difficult to overcome. In the district court, new evidence is reviewed de novo.
  • Money & Expertise: Both appeals to the Federal Circuit and civil actions in the district court can be quite expensive relative to ordinary prosecution costs. In addition, because most patent prosecutors have only limited federal court experience, a litigation or federal circuit specialist would likely need to be brought onboard the team.
  • Alternative Approach: The courts are typically asked to answer a yes/no question on claim validity. However, patent applicants recognize that a patent with slightly narrower scope (through amendment) may be almost just as valuable as the original broad claims. With that in mind, many applicants return to the original examiner with proposed claim amendments that lead quickly to allowance.

What other reasons explain the low rate of using the courts to challenge USPTO decisions.

33 thoughts on “Challenging USPTO Decisions in the Courts

  1. 33

    DC, one possible reason why the $25K spent on prosecution may not correlate that well with the potential value of the patent is the time it takes to prosecute the patent.

    For example: when the patent was filed, it was considered to have value. The prosecution then continued on “cruise control” mode (i.e., the standard prosecution strategy for “high value” patents”) till the BPAI decision. At that point, the in-house took a “second look” at the patent and decided that it is no longer worth pursuing. A common scenario: the patent may have been in a “hot” technology 4 years ago, but that technology area may have cooled down considerably since then.

  2. 32

    In 6’s eyes the “smearing” has nothing to do with reality and everything to do with the exposing of his man-crush Duffas fiasco (no doubt, the HMS Boundy and Tafas are also “tools” in his eyes, who only are out ta spread their negative opinions).

    Worse than clueless.

  3. 31


    assuming everything to be equal in a zero sum game

    What in the world would possess you to make such an unbridled and unsubstantiated assumption?

  4. 29

    I organize these by art unit every morning for an hour, been doing it for a year and a half and I link to it above because I’m duh winning! I’m putting them in a spreadsheet I link to on the left, I only have about 2/3rds of them, but my goal is to go back and get at least the examiner’s name and a link to the opinion for all of them. In particular Art Units say 1700 Chemical and Mechanical Engineering the same Examiner dukes it out with the same company say Dow or Bayer over and over sometimes affirmed, sometimes reversed. Sometimes the Judges will slap down an examiner in two different decisions in one day. I just look for relevant case law and I quote it when I see it, but come on, ping’s comment not withstanding, on the whole the Administrative Judges do a great job…

  5. 28

    What’s got you peasants all rilled up now? Jes us christ. I had a little fun with a newbie before I saw D set him straight. Am I like the dev il now?

    Or are you concerned with my pointing out Ron actively attempting to make himself a toolbag? If he’ll stop his actively attempting to make himself a toolbag and constantly whining then I will lay off him. Be clear, he is not here to be constructive in his criticism. He’s just here to smear the PTO at any opportunity, though he has a special affinity for appeals, and spread his negative opinion. Get him to curb his nonsense to where it is remotely balanced with what we all know is happening irl and I might even consider laying off him then. And have him present his reasonable big solution to whatever problem he is currently bit ching about. Is that a fair deal? I think it is more than fair.

    Same response to you lurker. And stop lurking btw. Makes you seem like a CAFC member or something.

  6. 27

    Here is a similar story

    In a notice of proposed rulemaking published in today’s Federal Register (76 Fed. Reg. 18990), the U.S. Patent and Trademark Office announced a proposal to revise the patent term adjustment (PTA) and patent term extension (PTE) provisions of the patent rules. In particular, the Office is proposing changes to the PTA and PTE provisions with respect to some circumstances of appellate review and for the submission of certain information disclosure statements (IDSs).

  7. 26

    Wow 6, your comments on this post do not reflect well on you. To the degree you care about your power to persuade through commenting, that power has diminished in my eyes.

  8. 25

    Ping, assuming everything to be equal in a zero sum game the average win rate should be 50%. But things aren’t equal. The examiner has a day or less to prepare a response to the brief, and the average examiner does not have the same legal training and legal writing skills as the average member of the patent bar. I’d expect arguments prepared by attorneys who are assumed to have better persuasive writing skills and who have much more time to prepare their brief to have a much better per-argument win rate than 39%, even if the sample of arguments that reach a decision is skewed by “cherry-picking” of cases.

  9. 24

    6 youre a scourge on this blog. A bright guy no doubt, but utterly lacking in posting tact or discretion. That you have not been banned demonstrates Dennis’ infinite patience.

  10. 23

    Even cherry-picking, the best the Office can do is 61%.

    That is sad.

    I wonder what the number would be if every case appealed had to go to the Board (get rid of the mulligan kick-backs).

  11. 22

    “I’m guessing that if we had rejection-by-rejection metrics, we would see an affirmance rate well below 50%”

    If you read Dennis’ article you’ll see that your guess is incorrect. Issue-by-Issue analysis resulted in a 61% affirmance rate.

    I personally found the article very informative.

  12. 19

    I hear the cry of “Me too me too wait for me” from 6 trying ta keep up with his posse by making fun of the less well initiated. Wouldn’t it be a kick if Sakura Kirchen was a cancer survivor as well?

  13. 18

    Nice 6 – callin the one dude who has not had any problem putting substance behind his arguments on most any blog space a “tool”


  14. 17

    as I expect many of these to show substantial improvement and subsequent decline in the number of appeals.

    Not likely – have ya seen the RCE lag? Expect more, not less ta go the way of Appeal in order ta force any substantive negotiations.

  15. 16

    Dear Dennis,
    Regarding the comment and reply of April 9, 3:42 and 4:11, it would be interesting to know what is the average cost of a patent that does not go to appeal, the average cost of going to appeal at the board of appeals, and the average cost of goint to appeal at the CAFC or other district courts. It is difficult to assess the significance of the $25,000 without these other numbers. Have you posted these statistics in prior blogs?
    For example, it seems to me that the cost of filing a case at the CAFC could be in the $100,000 ball park (at least I know of some cases that cost that much), whereas going to appeal to the Board of Appeals is more likely to be in the $6000 ball park (and is really unlikely to be anything close to $100,000, but I don’t have reliable statistics for those numbers either), and just because the Applicant already spent $25,000 on the entire case doesn’t mean they are willing to spend another $100,000, despite any correlation between the $25000 figure and how important the Applicant feels the case is. Prehaps this is essentially the point of the prior comment, but my point is just that the actual dollar amounts makes all the difference in the world as to the significance of the comment. However, I don’t fully agree with the sunken cost approach, even if the dollar amount of a filing an appeal at the CAFC is likely to be $100,000. Although the sunken cost approach may be basically correct, I suspect that there are many applicants that are not quite as sophisticated in their analysis, especially amongst the small time inventors, and just think in terms of what will going to appeal at the CAFC cost (which may be completely unaffordable to a small startup)compared to what will it cost to amend the case and refile, without a significant analysis of the breadth of the claims that will result and the value of the added breadth (which would seem to be needed for the sunken cost approach)- especially considering that there would seem to be a significant amount of uncertainty in the breadth of the resulting claims as a result of uncertainties over how a court would interpret the claims during a subsequent infringement suite, after the claims are read in light of the specification and prosecution history.

  16. 15

    Dennis, my point is that virtually all filers of appeals, including those whose appeals the Office selects for BPAI review, file an appeal in order to have substantive negotiations on the contours of their claims. Even a Board decision to affirm the examiner is only a step in this negotiation process – leading in most cases to further claim amendments to capture allowable subject matter. This is not about the “quality” of the case – a notion you may be importing from judicial review of district court decisions, where the legal position of the parties is fixed and where there is no opportunity to amend the claims. A decision of an applicant not to challenge BPAI decisions in court is by no means indicative of his “low quality” position. It is often the case that the applicant’s position is merely that the examiner had not met the burden of showing unpatentability. Once a substantive response is obtained, that applicant’s position disappears regardless of the merits of the examiner’s response. Moreover, many board affirmances reject claims on grounds other than those advanced by the examiner and thus mere affirmance does not prove anything regarding the “quality” of the applicant’s position. Rather, the low BPAI challenge rates simply indicate that unlike district court decisions, the applicant has other lower risk options for securing his patent rights.

    I should also point out that the statistics in your 2009 study come from cases that were filed well before Director Kappos took his position and made substantial improvements in the examination process. I encourage you to continue monitoring the broader appeal statistics as I expect many of these to show substantial improvement and subsequent decline in the number of appeals.

  17. 14

    Could it be, Dennis, that the one of the reasons for so few appeals is that the Board explains it reasoning so well that most applicants will accept an adverse decision?

  18. 12

    Toolbag! It’s been too long! Where have you been man? Where were you on the Jung case posting? Work a burden on you lately? Iirc you stood proudly on the losing side of that one! Maybe you could say a word on that in retrospect?

    In any case, D’s analysis is not confusing for anyone with any experience in patent matters. Take your toolbag whiner attitude someplace else.

    “Is right?”

    Is of course right D. and that’s all anyone with a brain took your original analysis to mean. Toolbag just wanted a reason to whine.

  19. 11

    This is true with almost any appeal in any area of law. The party that lost at the lower tribunal typically has a right to appeal. However, the party that won below can (almost) always end the appeal by forfeiting the case.

  20. 10

    Ron – Thanks for this post. I agree with you that substantive examination commonly requires at least three rounds with the patent examiner — that means that the applicant will either need to appeal or file an RCE to get to the point. As you imply, many appealed cases do not make it to a BPAI decision. Rather, after finally considering the applicant’s substantive arguments, the PTO withdraws its rejection and re-opens prosecution. It appears that the withdraw/re-open process is not random, but rather that the PTO is more likely to withdraw its rejection when it does not have a good case on appeal.

    Still, some number of appeals make it through to the BPAI and the BPAI decided those cases and the vast majority of applicants accept the BPAI decision rather than pursuing further challenge. You don’t say this, but your ‘cherry picking’ analogy suggests that the applicant’s who actually receive a BPAI decision probably have lower-quality arguments. Is right?

  21. 9

    Is it true that the USPTO has discretion over which appeals will proceed to the Fed Cir? I’ve heard that the Office of the Solicitor will re-open or allow cases even following BPAI affirmance that it doesn’t want to defend at the Fed Cir. Is this true?

  22. 8

    But that is only a soft version of the sunk cost fallacy. To take an analogy, if you invest $25k in a company and see it go down the tubes, the same theory would say that your initial investment “probably correlates with the potential value” of the company, so you should invest even more.

  23. 7

    Although I know it was not your intention and there is some value in your analysis, it is biased and incomplete because implying that BPAI statistics represent actual appeal statistics is misleading. During the period you were looking at in the study you posted on SSRN, more than 25,000 appeal briefs were filed per year but only about 7,000 appeals per year actually reached the BPAI. You failed to study a fair sample of the full ensemble of appeal briefs filed. This may lead to your bewilderment and the question you propounded for the readers, as you overlooked a real possible explanation for the low rate of attempted judicial review of BPAI decisions: It is that the vast majority of appeal briefs are filed as mere attempts at receiving substantive examination – not as a legal challenge to well-considered agency action. This is particularly true in cases where office actions before final rejections do not substantially advance prosecution – when the examiner fails to follow the sound instructions on examination in Chapter 2100 of the MPEP. Once real substantive exchange finally takes place (after an appeal brief is filed), prosecution often commences without having to resort to judicial review. As the number of appeal briefs filed per year has tripled, BPAI has only slightly increased its inflow. You have studied only those appeals that were cherry-picked by the Office to go to the BPAI and it is not surprising that as progressively smaller fraction of appeals are selected for BPAI review, you find in your research that “it also appears that the affirmance rate has increased over time … and that the rate of reversal has decreased.” The reason for this trend is easily seen by looking at the dramatically growing gap between the two curves of the appeal statistics shown in Figure 3 at link to .

    Your derived fraction of affirmances and reversals of appeals thus obtained are therefore arbitrarily biased and the temporal trend meaningless as it is a mere reflection of the progressively smaller share at the top of the appeals “pyramid” ranked and cherry-picked by the Office for BPAI review.

  24. 6

    TJ – I agree that we should be careful of the sunk cost fallacy, but I disagree that my use of the $25k in prior prosecution costs succumbs to that problem. My argument is that money spent on prosecution probably correlates with potential value of the patent. We could quibble with the strength of that correlation, but it is not the sunk cost fallacy. In particular, I’m not saying that they should pursue the the appeal because they have already spent money on the application.

  25. 5

    Your quoting the $25,000 figure is a classic instance of the sunk cost fallacy. The question is not whether the invention is worth $25,000, it is whether the marginal increase in claim coverage from winning an appeal–discounted by the chance of winning–is worth the marginal cost of the appeal. You note this in your second and fourth points, but they also show that the $25,000 is completely irrelevant.

  26. 4

    “the BPAI has been fairly good at granting rehearing requests when there is a clear and easily identifiable mistake in their opinion.”
    That is a big *&^%$ joke. The BPAI’s worst decisions are their Decisions on Request for Rehearing. The BPAI will go to great lengths to ignore case law when a “clear and easily identifiable mistake” has been pointed out.

    “Under Zurko, the Federal Circuit is required to give deference to the USPTO’s findings of facts and only overturn those findings when “unsupported by substantial evidence.” This standard is difficult to overcome.”
    Read enough BPAI decisions and you’ll see that it is a pretty easy burden to overcome. The BPAI’s findings of fact are often no more than “we agree with the Examiner – case closed.”

    “Both appeals to the Federal Circuit and civil actions in the district court can be quite expensive relative to ordinary prosecution costs.”
    Bingo – that is the real reason.

    “Sometimes the BPAI rejects the examiner reasoning but finds an alternative ground for rejection.”
    Then that application should not be affirmed. Under 37 CFR 41.50(a)(1), “[t]he Board, in its decision, may affirm or reverse the decision of the examiner in whole or in part on the grounds and on the claims specified by the examiner.” The rules do not permit the BPAI to affirm the decision of the examiner based on grounds specified by the BPAI. The BPAI can specify new grounds of rejection under 37 CFR 41.50(b), but they cannot affirm the Examiner when the BPAI is not relying upon the grounds specified by the examiner.

  27. 3

    Dennis, re your bottom line question, you have covered most of the answers I am aware of, especially the costs, the low odds of appellate success [other than for interferences], and the available option of further PTO prosecution in a continuation or divisonal.
    One other is that if the application is considered important enough to consult an experienced Fed. Cir. practioner about appealing from the Board to the Fed. Cir., they seem to quite often advise the client to first file a continuation to get professionally drafted supporting declaration evidence into the record, to improve the odds of success on appeal.

  28. 2

    Right – by “entirely affirmed,” I really mean that all of the claims that were rejected by the examiner still stand rejected at the end of the BPAI appeal. Sometimes the BPAI rejects the examiner reasoning but finds an alternative ground for rejection.

    In an article a couple of years ago, I looked at rejection-by-rejection results: link to

  29. 1

    “In about half of those decisions, the examiner rejections were entirely affirmed.”

    DC, I would note that several opinions bearing the term “affirmed” at the bottom include reversed rejections – just all appealed claims still had some affirmed ground of rejection. That is, the Board uses “affirmed-in-part” only when a claim has all rejections on the claim reversed. So I’m curious if your “entirely affirmed” metric (like the USPTO’s) is based on simply the decisions bearing the affirmed title or if it truly accounts for each rejection considered and affirmed on appeal.

    I’m guessing that if we had rejection-by-rejection metrics, we would see an affirmance rate well below 50%. Ron K?

Comments are closed.