GLP lab notebooks — how they differ from an ordinary research notebook

Search "GLP lab notebook" and you'll find plenty of general explanations of what GLP (Good Laboratory Practice) is, but very little on how it actually differs from the "National R&D Research Note Guideline" we've covered elsewhere on this blog. Treating the two as the same thing works fine — until the moment you start a nonclinical study for a drug, medical device, or cosmetic, and run into an entirely different legal basis and vocabulary. This post lays out that difference.

GLP comes from a different legal system entirely

Most of what this blog has covered as "research notebook" rules rests on Korea's National R&D Innovation Act and its subordinate directive, the National R&D Research Note Guideline. It applies to institutions performing government-funded R&D projects.

GLP (비임상시험관리기준, Good Laboratory Practice) has a completely different legal basis: Article 34-3 of the Pharmaceutical Affairs Act, Article 37 of its enforcement rule (Regulation on the Safety of Pharmaceuticals, etc.), and the notification issued under that delegation — the Standards for Nonclinical Laboratory Studies (a Ministry of Food and Drug Safety notice, most recently amended as Notice No. 2022-93, effective January 1, 2023). Applicability has nothing to do with whether the project is government-funded — any institution running nonclinical safety and efficacy studies for pharmaceuticals, quasi-drugs, cosmetics, medical devices, or health functional foods must be designated as a nonclinical testing institution and comply, government money involved or not, private company or not. The regulator is different too: the Ministry of Food and Drug Safety, not the Ministry of Science and ICT. Always confirm the exact provisions against the Pharmaceutical Affairs Act and the Standards for Nonclinical Laboratory Studies themselves.

It's not called a "research notebook" — it's called "raw data"

The terminology itself differs. The National R&D guideline builds its rules around a single document unit called the "research notebook." GLP instead uses the broader concept of "raw data": not just a paper lab notebook, but instrument printouts, photographs, electronic records, and microfilm copies — any first-generation record produced during the study, regardless of format.

Where raw data is handled electronically, GLP's central requirement is the audit trail. Every time data is changed, the reason for the change, the date and time, and the signature of who made it must all be recorded alongside it — and the original value must remain visible, never overwritten. The underlying principle is the same as the National R&D guideline's requirement to detect tampering (see our post on what an electronic research notebook actually needs), but GLP is far more specific about exactly what the audit trail has to capture.

The two systems, side by side

The line that trips people up most in practice is custody. Under the National R&D guideline, it's enough for a researcher to hand notebooks over to a department the institution designates. GLP goes further, with a distinct concept — the "archive facility" — and requires that facility and its archivist to be formally designated. Leaving raw data in a general document room or on a personal computer satisfies neither rule, but GLP is noticeably stricter about confirming exactly where, and under whose responsibility, it's kept.

What's easy to miss when both regimes apply at once

Nonclinical studies run as part of a government-funded new-drug or medical-device development project — a common setup — are subject to both regimes simultaneously. It's easy to assume that meeting the National R&D Research Note Guideline is enough, only to discover during an MFDS inspection that GLP's raw-data, audit-trail, and archive-facility requirements are checked separately. Satisfying one does not automatically satisfy the other.

The bar for adopting an electronic system also differs. An ordinary electronic research notebook needs three things — signature authentication, automatic timestamping, tamper detection. A GLP environment often requires, on top of that, formal system validation. Skip that validation step when bringing an ELN or LIMS into a GLP setting, and the system can work perfectly well while still not being recognized as compliant.

Self-check

• [ ] Have we confirmed whether the studies we run fall under the

nonclinical testing institution designation requirement?

• [ ] Does our internal policy define the scope of "raw data" (including

electronic records, instrument printouts, and photographs)?

• [ ] Does our computerized system actually produce an audit trail

(change reason, timestamp, signature, and the preserved original value)?

• [ ] Are a dedicated archive facility and archivist formally designated?
• [ ] If a project is subject to both the National R&D guideline and GLP,

are both sets of requirements being checked separately?

• [ ] If we introduced an electronic system into a GLP environment, did it

go through system validation?

The last two items are the ones most often skipped. Both regimes share the same underlying goal — nothing can be altered after the fact — but GLP spells out separate procedures for reaching it (archive-facility designation, system validation). This post is a general overview; always confirm actual applicability and the exact provisions against the Standards for Nonclinical Laboratory Studies and Ministry of Food and Drug Safety guidance.