nanalLabs BlogSearch "research notebook audit" and the top results are usually an institution's own routine inspection checklist or a lost-notebook report form. In practice, two outside bodies actually look at research notebooks — the specialized agency or ministry overseeing government R&D projects, and the National Tax Service, which reviews R&D tax credit claims after the fact. This post lays out what each one checks and on what legal basis, plus the single most common finding in practice: backdating.
Two different audits, not one
The two audits differ in legal basis and in what they're actually verifying.
| Government R&D oversight audit | Tax credit post-filing review (NTS) | |
|---|---|---|
| Legal basis | National R&D Innovation Act and its enforcement decree (full text) | Enforcement Decree of the Restriction of Special Taxation Act, Article 9 |
| Conducted by | Specialized agency, responsible ministry, or the institution's own audit office | NTS headquarters and regional tax offices |
| What it checks | Integrity of R&D spending and project execution | Whether the tax credit's requirements were actually met |
| If flagged | Participation restrictions, clawback of funding, research-misconduct proceedings | Credit clawback plus penalties |
We've already covered corporate research institutes' tax-credit basis and self-inspection items in our post on the corporate lab notebook, and an institution's own routine inspection in research notebook inspection. This post focuses on the two outside audits themselves. One concrete example: a National Institute of Health regular comprehensive audit report noted that while the institute had an incentive system rewarding good research notebooks, it had no corresponding sanction for researchers who simply didn't keep one — and flagged that gap as needing a fix (audit report). The finding wasn't about any individual notebook — it was about the absence of a system to enforce keeping one.
What an audit actually looks at, in three stages
- Document review — cross-checking whether the notebook's content
matches the project's own plan, report, and settlement documents. A mismatch in dates, participants, or experiment details is itself grounds for a first-round finding.
- Physical verification — checking not just the notebook but physical
outputs like prototypes or equipment. Deliverables that exist only on paper are usually the first thing an audit catches.
- Interviews — asking participating researchers directly whether they
actually ran the experiment and can explain the procedure written in the notebook. If the person who signed it didn't write it, or can't explain it, the signature alone won't save its credibility.
The most common finding — backdating
The most frequently cited finding in practice is backdating: not writing in the notebook as work happens, then filling it in all at once after an audit or tax investigation notice arrives. For paper notebooks, forensic analysis of handwriting, ink oxidation, and paper wear can expose a gap between the date written down and when the entry was actually made. Cloud documents have the opposite problem — without a preserved edit history, there's often no way to prove when something was written at all, which makes the suspicion hard to clear. Since 2020, the NTS has run a tax credit pre-review program that encourages companies to have their supporting evidence — including research notebooks — checked before filing. And enforcement appears to be tightening: reporting suggests the NTS has been stepping up post-filing verification of R&D tax credits in recent years (coverage here) — for exact year-by-year figures on findings and clawbacks, check the NTS's own published data directly. If you want to defend against a backdating claim, the timestamp needs to come from an external, third-party system — not just a date written inside the notebook itself.
If you've received an audit notice, check this now
- [ ] Do the notebook's dates, participants, and content match the
project plan, report, and settlement records?
- [ ] Can you produce the physical prototypes, equipment, or outputs the
notebook references?
- [ ] Can the researcher who wrote the notebook explain the procedure in
an interview?
- [ ] Can you prove when an entry was actually written using something
other than the notebook itself?
- [ ] Do corrections follow the required method (strikethrough plus
signature, no rewritten pages)?
Bottom line
Unlike an institution's internal inspection, an outside audit arrives without warning and combines document review, physical verification, and interviews. Both kinds of audit ultimately converge on the same question: how do you prove a record was actually written when it says it was? Always check the current statutory text and NTS notices directly before relying on any of this.