Secret Clearance Denial Rate: Why Level Alone Does Not Predict an Outcome

Posted by Ashley Jones

A Secret clearance denial rate does not exist as one published number. The statute that calls for a denial percentage does not ask for it by clearance level, and the fiscal 2017 federal report that printed agency figures showed ten agencies ranging from 0.0 percent to 5.9 percent in a single year. Any page that quotes one Secret figure owes you four answers: denials or revocations, out of what, over which years, and for whose cases.

That matters because the number gets used two ways. It is used to scare (“one in so many gets denied”) and to reassure (“almost nobody is denied”). Neither use survives a look at what the number actually counts.

Key takeaways

  • The denial-or-revocation percentage in 50 U.S.C. 3104(b)(2)(G) is one figure across initial and periodic investigations, with no clearance-level split in its wording.
  • In the ODNI report for fiscal 2017, ten unnamed agencies had denial percentages from 0.0 to 5.9 percent, and the report ties the gap to how each agency processes cases.
  • DCSA reported 1,810 cases denied or revoked against 836,461 total cases adjudicated in fiscal 2021. That is about 0.2 percent on a mixed base, not an applicant rate.
  • GAO found 63 percent of the 305 clearance statistics it checked for one quarter of fiscal 2024 inaccurate or incomplete.
  • SEAD 4 tells adjudicators to judge each case on its own merits, using nine listed factors. Clearance level is not one of them.
For employers
Hiring cleared professionals?
Post your cleared req where security-cleared candidates already search.

Post a Cleared Job

For job seekers
Holding a clearance?
Browse cleared roles from employers hiring right now.

Browse Cleared Jobs

Is there a single Secret clearance denial rate?

No. The statute asks for counts of clearance holders and approvals by level, and for a denial-or-revocation percentage as one figure. A published percentage describes an agency, a year, and a counting method. It does not give the odds for the next Secret applicant.

Section 3104(a)(1) of Title 50 asks the President’s annual report to count, for each clearance level, how many people held a clearance and how many were approved. Those are holder and approval counts, not denial counts. The denial percentage sits in subsection (b)(2)(G), the intelligence community report, and reads as one percentage of investigations, “including initial and periodic” reinvestigations. The 2023 edition of the Code on govinfo shows the section as last amended in 2019.

The criteria that govern individual outcomes are the same at every level. SEAD 4, effective 08 June 2017, sets “the single, common adjudicative criteria” for everyone who needs eligibility for access to classified information. Its list of factors an adjudicator considers has nine entries, from the seriousness of the conduct to the likelihood it will recur. The requested clearance level is not on the list.

Does a denial rate count denials, revocations, or both?

It depends on the report, and mixing the two changes the meaning. ODNI’s fiscal 2017 table split them: denials from initial cases, revocations from reinvestigations or revocations for cause. DCSA’s fiscal 2021 report printed one pooled count.

In the ODNI report, Table 5 gives each of ten agencies a denial percentage and a revocation percentage. Agency #1 shows 5.6 percent denials and 1.4 percent revocations. Agency #5 shows 0.2 percent and 0.6 percent, so its revocation figure is the larger one. A reader who adds the columns, or who quotes only the larger one, has built a number the report never printed.

The DCSA fiscal 2021 year-in-review prints “Denied Revoked” as one figure, 1,810. The same page lists about 16,380 preliminary letters of intent to deny or revoke and about 1,744 final letters. Even the word “denial” has more than one stage to count.

What is the denominator behind the percentage?

Often it is not stated. A percentage without a base is a ratio of unknown things. When the base is stated, check whether it holds only the cases the headline is about.

The statute itself uses two bases. Subsection (b)(2)(C) counts initial investigations “adjudicated with notice of a determination,” favorable and unfavorable. Subsection (b)(2)(G) takes a percentage of “security clearance investigations.” Those are different units, and a reader quoting “the denial rate” has to say which one.

The DCSA page that prints 1,810 describes the work behind it as “personnel security, suitability, and credentialing actions,” and its list of requests received includes continuous evaluation (42,348) and incident reports (31,675) next to initial investigations and reinvestigations. Dividing 1,810 by 836,461 gives about 0.2 percent. That is plain division on a mixed base, and the report does not present it as a rate for first-time applicants.

Are Confidential and Secret counted together?

Often, yes. For the government-wide count of clearance holders and approvals, the statute says the President “may consider” Confidential and Secret as one level, and Top Secret and above as another. ODNI’s fiscal 2017 tables print Confidential and Secret together, so a figure with “Secret” attached may describe both lower levels or no level at all.

The fiscal 2017 ODNI report labels its holder and approval tables with a “Conf/Secret” column beside “Top Secret,” and its timeliness table uses a “Secret/Confidential” column. Its denial table has no level columns at all. A writer who turns these reports into a Secret-only denial figure is attaching a label the source does not carry.

Why can one agency’s rate not be applied to another agency’s applicants?

Because agencies count differently and serve different people. ODNI said the gap in denial percentages “can be attributed to the various processes employed by those agencies.” Applying one agency’s percentage to a different pool of applicants carries that difference over.

The ODNI report gives two process examples. Some agencies “may discontinue security processing due to automatic disqualifiers found during a suitability for federal employment review,” and those cases are not categorized as clearance denials. Other agencies “consider all relevant information” and “render security clearance denials based upon the totality of the information contained in the case files, which results in a higher percentage of denials.” In that one year, within one government’s intelligence community, denial percentages ran from 0.0 to 5.9 percent.

Populations differ too. ODNI sent its fiscal 2017 data call to sixteen intelligence community elements, and Table 5 lists ten agencies. Table 5 does not split denials between government employees and contractors, although the report’s Table 7 lists contractor cases at the same ten agencies. It describes those agencies’ caseloads, not the customer mix in DCSA’s report. DCSA’s report lists its customers: Army 30 percent, Navy 29 percent, industry 24 percent, Air Force 14 percent, and the Fourth Estate 2 percent. For the industry share, 32 CFR 117.10 says the contractor “must determine that access to classified information is essential” for the work, will “limit requests” to the minimum number of employees necessary, and will not submit requests for people who are not its employees or consultants. That paragraph binds the cleared company, not the applicant. It does mean contractor requests come from a group an employer already selected, so a rate computed on them describes that group. ClearedJobs.NET’s read of the fiscal 2014 ODNI report reached the same caution about agency-to-agency comparisons.

Source What it counts What sits under the percentage What it cannot tell a Secret applicant
ODNI report, fiscal 2017, Table 5 Denials (initial cases) and revocations (reinvestigations or revocations for cause), as separate percentages for ten unnamed agencies Percentages only; no base is printed in the table, and agencies run different processes Which agencies they are, and anything by clearance level
DCSA year-in-review, fiscal 2021 1,810 “Denied Revoked,” pooled 836,461 total cases adjudicated, within personnel security, suitability, and credentialing work A first-time applicant rate
GAO-26-107100, December 2025 Not a rate: a check of 305 statistics from seven agencies, third quarter of fiscal 2024 Eight of the 39 reporting requirements ODNI sets The error rate of all federal clearance data

Do completed-case rates include withdrawn, discontinued, or pending cases?

Not necessarily. In ODNI’s fiscal 2017 report, some agencies’ cancelled cases are not counted as denials, and pending and incomplete cases are reported on their own lines. A rate built on completed cases leaves them out.

The ODNI report says some cancelled cases “are not categorized as security clearance denials.” The statute separates three more groups from denials: pending investigations in subsection (b)(2)(E), investigations that ended in incomplete information in (H), and investigations that did not produce enough information to decide on potentially adverse information in (I). In fiscal 2017, Agency #9 reported 1.8 percent of its investigations as incomplete and Agency #4 reported 6.7 percent. Each sits on its own line, apart from the denial table.

Then there is the question of whether the completed-case counts themselves are sound. In GAO-26-107100, published Dec 11, 2025, GAO examined 305 statistics ODNI collects and found 191, or 63 percent, inaccurate or incomplete. Its “completed adjudications” example: one agency reported 839 Top Secret contractor clearances in a quarter, and GAO’s case-level review found 142. Agency officials also told GAO they had included continuous vetting adjudications in that figure, which differed from ODNI’s instructions.

GAO looked at seven agencies and eight of 39 reporting requirements, and calls its observations not generalizable. The seven agencies do report for over 90 percent of people holding a clearance, but 63 percent is not the error rate of denial figures specifically.

What decides an individual outcome?

The adjudicator’s weighing of the facts of that case. SEAD 4 says each case “must be judged on its own merits,” and that adjudicators consider all available, reliable information about the person, favorable and unfavorable.

The directive keeps both halves of the rule in view. Eligibility “shall only be granted” when the evaluation “demonstrates that such eligibility is clearly consistent with the interests of the United States; any doubt shall be resolved in favor of the national security.” It also says one criterion may not be enough for an unfavorable decision, though a recent or recurring pattern can be, and that “a single criterion may be sufficient” even without one.

The investigation behind a case is not the same at every level. Under 32 CFR 117.10(b)(1), one investigative tier covers Confidential and Secret positions and another covers Top Secret and SCI, so what an adjudicator sees can differ by level even though the criteria do not. That is one reason a Secret-versus-Top Secret comparison needs its own source.

The thirteen guidelines, A through M, are each “evaluated in the context of the whole person.” That is the weighing described in ClearedJobs.NET’s explainer on the whole-person concept. SEAD 4 also describes two kinds of conditional approval. Where information is serious enough to warrant a denial recommendation but the risk “can be managed with appropriate mitigation measures,” the adjudicator may recommend approval with an exception. Where it is not serious enough for a denial recommendation, the adjudicator may recommend approval with a warning that future incidents may result in revocation. Other things sit on the same scale: how recent the conduct is, what the person did about it, and whether the answers on the SF-86 were complete. SEAD 4 calls intentional material falsification “of significant concern.”

For a particular concern on your own file, the deciding office is the adjudicating agency. For a contractor employee, 32 CFR 117.10 says the cognizant security agency determines eligibility under SEAD 4 and notifies the employer of a denial, suspension, or revocation. Questions about your own file go to the adjudicating agency’s security office or, if an employer sponsors your clearance, to that employer, which is the party the agency notifies. ClearedJobs.NET’s post on denial rates by adjudicative guideline is a related read on how published denials are broken down.

What does a responsible citation of a denial rate contain?

Five items: the period, the scope, the source, the numerator, and the denominator. A figure missing any of them cannot be checked, and a figure that cannot be checked should not shape a decision about your career.

  • Period. The fiscal year or quarter, as the source states it.
  • Scope. Which agency or agencies, which population (government employees, contractors, intelligence community elements), and whether levels are grouped.
  • Source. The report, its date, and the table or page, linked to the document and not to someone else’s summary.
  • Numerator. Denials, revocations, or both, and whether preliminary letters are counted.
  • Denominator. Investigations or adjudications, initial cases or all, and whether suitability, continuous evaluation, withdrawn, or pending work is inside it.

Apply this to the next denial percentage you read. Write the five lines under it. If the page cannot fill them, set the number aside and use the primary source it came from, such as the ODNI report or the DCSA year-in-review linked above.

ClearedJobs.NET connects cleared talent with the employers who need it.
Whether you are hiring for a cleared req or holding a clearance and looking, start here.

Post a Cleared JobBrowse Cleared Jobs

Frequently Asked Questions

What percentage of security clearances are denied?

No single government-wide percentage was found in the sources read. The fiscal 2017 ODNI report gave ten agencies denial percentages from 0.0 to 5.9 percent, and DCSA’s fiscal 2021 report printed 1,810 denied or revoked cases against 836,461 adjudicated, on a mixed base.

Is the denial rate higher for Secret than for Top Secret?

The sources read give no rate for either. The statute’s denial percentage has no level split in its wording, the ODNI denial table has no level columns, and ODNI groups Confidential with Secret in its other tables. DCSA’s April 2022 clearance working group update for the National Industrial Security Program advisory committee printed counts of denied or revoked industry cases by level (fiscal 2022 to date: 168 Secret, 26 Top Secret, 4 Top Secret/SCI), but no per-level base, so they cannot be turned into a Secret-versus-Top Secret rate. A claim of a difference needs its own source.

Does a Secret clearance denial rate predict my result?

No. SEAD 4 says each case is judged on its own merits, using nine factors about the conduct and the person. A rate describes a group of past cases under one agency’s counting method.

Why do different websites quote different denial rates?

They are usually reading different things. One may count revocations, another denials only, another pooled cases. Bases differ, years differ, and agencies process cases differently, as ODNI’s own report explains.

Are withdrawn or cancelled applications counted as denials?

The ODNI fiscal 2017 report says some cases cancelled before a determination are not categorized as clearance denials. Pending cases and investigations with incomplete information are reported separately in the statute.

Author

  • Ashley Jones is ClearedJobs.Net's blog Editor and a cleared job search expert, dedicated to helping security-cleared job seekers and employers navigate job search and recruitment challenges. With in-depth experience assisting cleared job seekers and transitioning military personnel at in-person and virtual Cleared Job Fairs and military base hiring events, Ashley has a deep understanding of the unique needs of the cleared community. She is also the Editor of ClearedJobs.Net's job search podcast, Security Cleared Jobs: Who's Hiring & How.

    View all posts

Comment

Notify me of updates to this conversation

Author

  • Ashley Jones is ClearedJobs.Net's blog Editor and a cleared job search expert, dedicated to helping security-cleared job seekers and employers navigate job search and recruitment challenges. With in-depth experience assisting cleared job seekers and transitioning military personnel at in-person and virtual Cleared Job Fairs and military base hiring events, Ashley has a deep understanding of the unique needs of the cleared community. She is also the Editor of ClearedJobs.Net's job search podcast, Security Cleared Jobs: Who's Hiring & How.

    View all posts
This entry was posted on Monday, November 16, 2026 11:36 am