Examples
Worked examples
- Is an instance
A contract research organization runs a 90-day repeat-dose rat toxicology study under a GLP-compliant protocol, with an independent Quality Assurance Unit auditing the raw data and final report before the Study Director signs it for inclusion in an IND's pharmacology/toxicology section.
- Is an instance
A university-affiliated preclinical testing facility conducts a genotoxicity assay following a written, approved SOP, with raw data, protocol deviations, and equipment calibration records retained in a controlled archive for later FDA inspection.
Counter-examples
Looks similar, but isn't
- Not an instance
An exploratory in vitro screening assay run to decide which of several candidate compounds to advance further is not a GLP study — it is discovery-stage research with no regulatory submission in view, and GLP obligations only attach once a nonclinical study is intended to support an FDA application.
- Not an instance
A student research project run with careful technique but without a Study Director, an independent Quality Assurance Unit, or SOP-driven documentation is not GLP-compliant — GLP is a formal regulatory quality system, not a general standard of careful benchwork.
Editorial commentary
Good Laboratory Practice (GLP) is the FDA quality-system regulation — 21 CFR Part 58 — that governs how nonclinical (preclinical) laboratory studies are planned, conducted, monitored, recorded, and reported when those studies are submitted to support an FDA research or marketing application: an Investigational New Drug (IND) application, a New Drug Application (NDA), a Biologics License Application (BLA), or a medical device submission. GLP does not apply to clinical trials in human participants, and it does not apply to manufacturing — it applies specifically to the animal and in vitro safety-testing studies that happen before a product ever reaches a human clinical trial.
GLP is one of three related but distinct FDA regulatory quality systems that a research administrator supporting translational or investigator-initiated work needs to be able to tell apart — informally known together as the “three GxPs.”
GLP vs. GMP vs. GCP: the three GxPs
Each GxP governs a different stage of a product’s path from the lab bench to a human clinical trial, and each has its own regulatory citation:
- GLP (Good Laboratory Practice) — 21 CFR Part 58 — governs the conduct of nonclinical (preclinical) safety studies: the animal toxicology, pharmacology, and safety-pharmacology studies that generate the data used to decide whether a compound is safe enough to test in humans in the first place.
- GMP (Good Manufacturing Practice) — 21 CFR Parts 210/211 (drugs), 21 CFR Part 820/the new Quality Management System Regulation (device), ICH Q7 (APIs) — governs how the investigational product itself is manufactured: facilities, batch records, process validation. See CASRAI’s Good Manufacturing Practice (GMP) guide for the full regulatory framework.
- GCP (Good Clinical Practice) — internationally harmonized as ICH E6 — governs the conduct of the clinical trial itself once the investigational product reaches human participants: informed consent, investigator responsibilities, data integrity. See CASRAI’s ICH GCP (Good Clinical Practice) dictionary entry.
Read in sequence, the three GxPs map directly onto a translational research program’s timeline: GLP-compliant nonclinical studies generate the safety data that supports moving a compound into human testing; GMP ensures the investigational product used in that testing is manufactured to a controlled, documented standard; GCP governs how the trial itself is then conducted once human participants are enrolled. A gap in any one of the three — not just GCP — can stall an IND submission or trigger an FDA inspection finding.
What GLP actually requires
21 CFR Part 58 applies to nonclinical laboratory studies of test articles regulated by the FDA — including food and color additives, animal food additives, human and animal drugs, medical devices, biological products, and electronic products — whenever those studies are intended to support a research or marketing permit application. The regulation’s core structural requirements include:
- A Study Director — a single named individual with overall responsibility for the technical conduct of the study and for the interpretation, analysis, documentation, and reporting of results. The Study Director is the single point of study control: they approve the protocol and any amendments, and confirm all data (including unanticipated test-system responses) is accurately recorded.
- An independent Quality Assurance Unit (QAU) — a function entirely separate from, and independent of, the personnel who direct or conduct the study. The QAU audits protocols, raw data, and the final report against the applicable standards, and reports its findings to facility management — a structural separation designed so quality oversight isn’t self-policed by the same team generating the data.
- A written, approved protocol for each study, plus documented standard operating procedures (SOPs) covering routine laboratory procedures.
- Calibrated, appropriately maintained equipment and defined test-system (animal, cell-based, or in vitro) husbandry/handling procedures.
- Complete raw-data retention and archiving — raw data, documentation, protocols, specimens, and the final report are transferred to a controlled archive during or at the close of the study, so the full record is reconstructable later by an FDA inspector.
These requirements exist to ensure that when FDA reviews a nonclinical safety study, the agency can trust that the data reflects what was actually observed — not what a sponsor hoped to see.
Why GLP matters for an IND or NDA submission
Under 21 CFR 312.23 (IND content and format), an IND application must include, for each nonclinical laboratory study subject to Part 58, a statement that the study was conducted in compliance with GLP — or, if it wasn’t, a brief statement explaining the noncompliance. In practice, FDA expects the pivotal nonclinical safety studies that establish a compound’s safety margin (the studies the IND review actually leans on to authorize human dosing) to be GLP-compliant; a pivotal safety study that isn’t GLP-compliant, or whose noncompliance isn’t adequately explained, is a real basis for an FDA clinical hold or a deficiency finding during review. Earlier-stage, purely exploratory nonclinical work (e.g., screening which of several candidate compounds to advance) is not required to be run under GLP — GLP obligations attach once a study is intended to support the regulatory submission, not to every nonclinical experiment a lab runs.
International recognition: OECD GLP and Mutual Acceptance of Data
Outside the U.S., the OECD Principles of Good Laboratory Practice serve the equivalent function and are structured very similarly to 21 CFR Part 58 (Study Director, independent quality assurance, SOPs, archiving). The OECD Principles underpin the OECD’s Mutual Acceptance of Data (MAD) system: a nonclinical safety study conducted under OECD Test Guidelines and OECD GLP Principles in one MAD-adherent country must be accepted for regulatory assessment purposes in every other adherent country, without being repeated. For a multinational translational program, this is what avoids having to re-run the same animal safety study separately for each regulator — provided the originating facility’s GLP compliance status (and its national GLP monitoring authority’s standing) is in order.
Frequently asked questions
Does GLP apply to a routine academic research lab?
Not to basic discovery-stage research. GLP obligations attach specifically to nonclinical safety studies that are intended to support an FDA research or marketing application — most commonly toxicology or safety-pharmacology studies feeding into an IND. A lab running exploratory assays to decide which compound to advance, with no regulatory submission in view yet, is not required to operate under Part 58.
Is GLP the same as GMP?
No. GLP (21 CFR Part 58) governs how nonclinical safety studies are conducted and documented. GMP (21 CFR Parts 210/211, among others) governs how the investigational product itself is manufactured. A program can be fully GLP-compliant on its toxicology studies and still have a GMP finding on its drug substance manufacturing, or vice versa — they’re separate quality systems assessed separately.
Is GLP the same as GCP?
No. GLP applies to nonclinical (animal/in vitro) safety studies conducted before human testing begins; GCP (ICH E6) applies to the clinical trial itself, once human participants are enrolled. Some institutions use “GLP” loosely to mean any lab that runs carefully, but the regulatory term specifically means Part 58 (or OECD GLP) compliance for a nonclinical safety study submitted to a regulator.
What happens if a pivotal nonclinical safety study isn’t GLP-compliant?
FDA requires IND applicants to disclose GLP compliance status for each nonclinical study subject to Part 58 and to explain any noncompliance. A pivotal safety study that is not GLP-compliant, without an adequate explanation, is a real risk factor for a clinical hold or a review deficiency — FDA is being asked to rely on data it cannot independently trust the integrity of.
Who has to run the Quality Assurance Unit — can the Study Director do it?
No. 21 CFR Part 58 requires the Quality Assurance Unit to be entirely separate from and independent of the personnel directing or conducting the study, including the Study Director. That structural independence is what gives the QAU’s audit findings credibility with FDA inspectors.
Related CASRAI resources
- Good Manufacturing Practice (GMP) — the manufacturing-quality counterpart to GLP
- ICH GCP (Good Clinical Practice) — the clinical-trial-conduct counterpart to GLP
- Clinical Trial Phases: FDA Definitions, Milestones, and Trial Administration
- Clinical Data Management: Processes, Systems, and Standards
- Clinical Research Administration — cluster hub
References
- 21 CFR Part 58 — Good Laboratory Practice for Nonclinical Laboratory Studies (eCFR)
- 21 CFR 312.23 — IND content and format (nonclinical GLP compliance statement)
- OECD Principles of Good Laboratory Practice and the Mutual Acceptance of Data (MAD) system
Also known as
GLP · 21 CFR Part 58 · Good Laboratory Practice for Nonclinical Laboratory Studies · OECD GLP · Preclinical GLP
Machine-readable encodings
Use in your systems
<role vocab="credit"
vocab-identifier="https://casrai.org/dictionary/"
vocab-term="Good Laboratory Practice (GLP)"
vocab-term-identifier="https://casrai.org/dictionary/term/good-laboratory-practice-glp" />{
"@context": "https://schema.org",
"@type": "DefinedTerm",
"@id": "https://casrai.org/dictionary/term/good-laboratory-practice-glp",
"name": "Good Laboratory Practice (GLP)",
"identifier": "https://casrai.org/dictionary/term/good-laboratory-practice-glp",
"description": "Good Laboratory Practice (GLP) is the FDA quality-system regulation (21 CFR Part 58) governing how nonclinical (preclinical) laboratory safety studies — chiefly animal toxicology and safety-pharmacology studies — are planned, conducted, monitored, recorded, and reported when submitted to support an IND, NDA, BLA, or other FDA research or marketing application; internationally, the OECD Principles of GLP serve the same function under the Mutual Acceptance of Data (MAD) system.",
"inDefinedTermSet": "https://casrai.org/dictionary/domain/compliance-regulatory#set",
"url": "https://casrai.org/dictionary/term/good-laboratory-practice-glp",
"sameAs": [
"GLP",
"21 CFR Part 58",
"Good Laboratory Practice for Nonclinical Laboratory Studies",
"OECD GLP",
"Preclinical GLP"
],
"license": "https://creativecommons.org/licenses/by/4.0/",
"publisher": {
"@id": "https://casrai.org/#organization"
},
"dateModified": "2026-07-17T05:08:45",
"inLanguage": "en"
}
