What Is Good Clinical Practice (GCP)?

Good Clinical Practice (GCP) is the international ethical, scientific, and quality standard for designing, conducting, recording, and reporting clinical trials that involve human participants. It is set out in the International Council for Harmonisation’s ICH E6 guideline and is incorporated, directly or by reference, into the regulatory frameworks of every major medicines regulator, including the FDA, EMA, MHRA, PMDA, and Health Canada. Compliance with GCP is not optional for regulated interventional trials — it is the baseline a sponsor, investigator, and institutional review board are expected to meet for the trial’s data to be considered reliable and its participants’ rights and safety to be considered protected.

This guide explains what GCP actually requires, who it applies to, how the standard has evolved from ICH E6(R1) through the current E6(R3) revision, how compliance is monitored and enforced, and how GCP relates to adjacent standards like GMP, GLP, and the training/certification landscape built around it.

The regulatory standard: ICH E6 and its revisions

GCP as a formal, harmonized standard originates with the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH), a body bringing together regulators and industry from the US, EU, Japan, and other participating jurisdictions. The foundational guideline, ICH E6, was first finalized in 1996. Two later revisions matter for anyone working in clinical research today:

• ICH E6(R2) (2016 addendum) added emphasis on risk-based quality management, sponsor oversight of outsourced trial functions, and electronic records/essential-document handling, without restructuring the original guideline.
• ICH E6(R3), the current core guideline, reached ICH Step 4 (final) on 6 January 2025. It restructures GCP around a set of principles plus annexes, sharpens the risk-based, quality-by-design approach, and explicitly addresses decentralized and pragmatic trial designs and electronic data sources that E6(R2) did not contemplate. EMA set 23 July 2025 as the effective date for the Principles and Annex 1 in the EU; FDA issued its own final E6(R3) guidance on 8 September 2025.

Because implementation dates differ by jurisdiction and by which annex applies, sponsors running multi-region trials need to track both revisions during the transition period rather than assuming a single global cutover date. For a full breakdown of what changed between the two revisions, see CASRAI’s guide to what changed under ICH E6(R3).

The core principles of GCP

ICH E6 sets out its requirements as a small number of core principles that everything else in the guideline operationalizes. In substance, these principles establish that:

• Trials must be conducted in accordance with ethical principles that have their origin in the Declaration of Helsinki and are consistent with GCP and applicable regulatory requirements.
• The anticipated benefits of a trial must justify the risks, established through a risk/benefit assessment before the trial begins and reassessed as it proceeds.
• The rights, safety, and well-being of trial participants are the most important considerations and must prevail over the interests of science and society.
• Available nonclinical and clinical information on the investigational product must be adequate to support the proposed trial.
• Trials must be scientifically sound and described in a clear, detailed protocol.
• A trial must be conducted in compliance with the protocol that has received prior IRB/IEC approval or favorable opinion.
• Medical care of, and medical decisions on behalf of, participants must always be the responsibility of a qualified physician or dentist.
• Everyone involved in conducting a trial must be qualified by education, training, and experience to perform their respective tasks.
• Freely given informed consent must be obtained from every participant before trial participation.
• All trial information must be recorded, handled, and stored in a way that permits accurate reporting, interpretation, and verification.
• The confidentiality of records that could identify participants must be protected.
• Investigational products must be manufactured, handled, and stored in accordance with Good Manufacturing Practice, and used in accordance with the approved protocol.
• Systems with procedures to assure the quality of every aspect of the trial must be implemented.

These principles are what a sponsor’s SOPs, a site’s Trial Master File, and a monitor’s visit checklist are all, ultimately, built to demonstrate compliance with.

For more detail, see new zealand — New Zealand requires formal pre-commencement regulatory approval for trials of new medicines, reviewed by Health Research Council committees on Medsafe’s behalf, running in parallel with mandatory HDEC ethics review.

The principles of ICH E6(R3)

The 13-point list above is the classic Section 2 framework from ICH E6(R2), which most GCP training materials and search results still reproduce. E6(R3), finalized at ICH Step 4 on 6 January 2025, replaced that structure: it restates GCP as 11 overarching principles, each broken into supporting numbered sub-points (1.1, 1.2, and so on) rather than a flat numbered list. Anyone comparing “the 13 principles of GCP” against a current E6(R3) training deck may notice the count and the wording have both changed — that is expected, not an error in either source. The 11 principles, drawn directly from the finalized guideline text, are:

1. Ethical conduct and participant protection. Trials must be conducted according to the ethical principles originating in the Declaration of Helsinki and consistent with GCP and applicable regulatory requirements; participants’ rights, safety, and well-being must prevail over the interests of science and society, and foreseeable risks must be weighed against anticipated benefits before a trial starts and as it proceeds.
2. Informed consent. Consent must be freely given, documented, and based on a process that leaves participants (or their legally acceptable representatives) genuinely well-informed — including provisions for emergency-consent situations and the use of technology to support the consent process.
3. Independent ethics review. Trials must be subject to independent review by an IRB/IEC, both before the protocol is implemented and periodically as the trial continues.
4. Scientific soundness. A trial must be scientifically sound for its intended purpose, based on adequate and current nonclinical and clinical knowledge, with that knowledge reviewed periodically as new information emerges.
5. Qualified individuals. Everyone involved — physicians, nurses, pharmacists, biostatisticians, monitors, coordinators, and others — must be qualified by education, training, and experience for their specific role.
6. Quality by design. Quality must be built into the trial’s scientific and operational design from the outset, with the factors critical to trial quality identified prospectively and strategies in place to prevent, detect, and address serious noncompliance.
7. Proportionality. Trial processes, monitoring approaches, and data-collection burden must be proportionate to the actual risks to participants and the importance of the data being collected — avoiding unnecessary complexity for its own sake.
8. A clear, feasible protocol. The trial must be described in a protocol (and supporting documents such as the statistical analysis plan and monitoring plan) that is clear, concise, scientifically sound, and operationally feasible.
9. Reliable results. Data-capture and computerized systems must be fit for purpose and proportionate to risk, essential records must be retained and available to regulators/auditors/IRBs on request, and trials must be transparently registered and their results publicly posted.
10. Clear roles and responsibilities. Sponsors may transfer, and investigators may delegate, specific activities, but each retains overall responsibility for what it delegates; agreements must document roles clearly, and appropriate oversight of delegated or transferred activities must be maintained.
11. GMP for investigational products. Investigational products must be manufactured under applicable Good Manufacturing Practice, retain their quality through handling and storage, and be used strictly in accordance with the protocol and relevant trial documents, including blinding requirements where applicable.

Source: ICH, E6(R3) Guideline for Good Clinical Practice, Step 4 final version, 6 January 2025 (database.ich.org). See CASRAI’s guide to what changed under ICH E6(R3) for the fuller picture of how the R2-to-R3 revision affects trial conduct beyond just this restructured principles list.

Who GCP applies to, and who is responsible for what

GCP assigns distinct, overlapping responsibilities rather than a single compliance owner:

• Sponsors are responsible for overall trial oversight, including selecting qualified investigators, implementing quality management and risk-based monitoring, ensuring safety reporting, and maintaining oversight of any functions delegated to a contract research organization (CRO) — delegation does not transfer the sponsor’s underlying regulatory responsibility.
• Investigators are responsible for the conduct of the trial at their site: protocol adherence, participant safety, accurate source documentation, obtaining informed consent, and appropriate delegation of tasks to qualified staff.
• IRBs/IECs are responsible for reviewing and approving the protocol, informed consent materials, and ongoing trial conduct to protect participant rights and welfare.
• CROs, where used, perform sponsor-delegated functions but operate under the sponsor’s oversight and within the sponsor’s documented quality system.

How GCP is incorporated into national and regional regulation

ICH E6 itself is a guideline, not a directly enforceable law in most jurisdictions — it becomes binding through national and regional implementation. In the United States, GCP-equivalent obligations are set out across FDA regulations including 21 CFR Parts 50 (informed consent), 54 (financial disclosure by clinical investigators, certified on Form FDA 3455), 56 (IRBs), and 312 (IND applications), with FDA’s E6(R3) guidance now incorporating the ICH text directly. In the European Union, GCP obligations are anchored in the EU Clinical Trials Regulation (EU) No 536/2014. Other ICH member and observer regulators — including the UK’s MHRA, Japan’s PMDA, and Health Canada — incorporate ICH E6 into their own clinical trial regulations and guidance in comparable ways. Medical device trials follow a related but distinct standard, ISO 14155 ("Clinical investigation of medical devices for human subjects — Good clinical practice"), rather than ICH E6, which was written for drug and biologic trials.

GCP training requirements

Regulators and funders increasingly require documented GCP training, not just GCP-consistent conduct. NIH policy (NOT-OD-16-148, effective 1 January 2017) requires all NIH-funded investigators and staff responsible for the conduct, management, or oversight of an NIH-funded clinical trial to complete GCP training, refreshed at least every three years, regardless of trial phase or intervention type. In practice, most sites and sponsors meet this through commercial or nonprofit training providers — CITI Program’s GCP course series (aligned to ICH E6, with separate tracks for drug/biologic and device trials) is one of the most widely used, and its completions are recognized under TransCelerate’s GCP Mutual Recognition Program across participating sponsor companies without separate retraining. This is distinct from the credentialing-exam route covered in our separate guide on GCP certification — GCP training documents that someone has completed a course; certification, where it exists, is a credentialing body’s exam-based credential built on top of that same underlying content.

How to get GCP-certified for free

A large share of searches for “GCP certification” are really asking one practical question: how to complete legitimate GCP training without paying for it. That is a reasonable thing to want, and it is worth being direct about it — something the commercial training vendors that dominate this search result have little incentive to volunteer. Two things are worth clarifying before comparing routes. First, none of the free or paid routes below issue a government-recognized license; every one of them produces either a course-completion certificate or, for ACRP/SOCRA specifically, a professional credentialing-exam credential built on top of the same ICH E6 content (see CASRAI’s dedicated guide to GCP certification for the full provider landscape, including the paid credentialing exams). Second, acceptance is never automatic — whether a specific sponsor, CRO, institution, or IRB will accept a given free course in place of its own required training is a decision made by whoever is imposing the requirement, not by the training provider.

Route	Cost	Who it’s realistically for	What you get
NIH/NIDA Clinical Trials Network GCP Training	Free, open to the public (including internationally)	Anyone wanting a genuinely no-cost, ICH-aligned course; not limited to NIDA-funded staff	A 12-module online course with per-module quizzes and a completion certificate valid for three years. It is not listed under TransCelerate’s GCP Mutual Recognition Program, so a specific sponsor may or may not accept it in place of its own required course — confirm first. See CASRAI’s NIDA GCP Training entry for more detail.
NIHR Learn (UK)	Free	People supporting clinical research delivery in the UK — NHS staff, UK university staff, and staff at other publicly funded organizations; you can register with a work or personal email	CPD-accredited GCP Introduction and Refresher e-learning, facilitated workshop versions, and a dedicated E6(R3)-aligned Investigational Medicinal Product management module. Built for and mainly recognized within the UK research-delivery system — don’t assume a non-UK sponsor will accept it without checking.
CITI Program	Often free in practice, but not free by default	Anyone affiliated with one of CITI’s roughly 3,000+ subscribing institutions (including an estimated 95% of Carnegie R1 research universities) — which is most academic researchers, whether or not they realize it	The training platform most U.S. academic institutions and hospitals already deploy for GCP. If your institution holds a CITI subscription, your access is already paid for — check with your research office before assuming you need to pay personally. CITI’s GCP courses are listed under TransCelerate’s GCP Mutual Recognition Program, which helps with cross-sponsor acceptance. Independent learners with no affiliated institution can register and pay per course directly, at a fee CITI or the institution sets (not centrally published).
TransCelerate GCP Mutual Recognition Program	Not a course — no direct cost	Anyone who has already completed a listed provider’s GCP course and needs it recognized by more than one sponsor company	Not itself a certification route: it’s an industry mechanism among participating pharmaceutical sponsor companies that lets a GCP completion from a recognized provider (CITI and certain others) count for multiple sponsors without separate retraining. It matters here because it’s often what actually determines whether a free course you’ve already completed will be accepted downstream. See CASRAI’s TransCelerate GCP Mutual Recognition Program entry.
ACRP (CCRC/CCRA/CPI/ACRP-CP) and SOCRA (CCRP)	Paid — exam and renewal fees apply	Clinical research professionals seeking a portable, career-level credential, not just a per-protocol training requirement	A proctored, experience-gated professional credentialing exam, not a training module. These are the routes most worth paying for once GCP training alone is not enough — see the full comparison, including renewal cycles, in CASRAI’s GCP certification guide.

The practical takeaway: if you just need documented GCP training and have no specific provider mandate from a sponsor or funder, check your own institution’s CITI subscription first, then NIH/NIDA’s free course if you have none, or NIHR Learn if you’re UK-based — before assuming a paid commercial course is the only option. Confirm acceptance with whoever is actually requiring the training before relying on any free route for a specific protocol.

How GCP compliance is monitored and enforced

GCP compliance is checked continuously during a trial, not just at its conclusion. Sponsors (directly or via a CRO) conduct routine monitoring visits — site initiation, interim, and close-out visits that include source data verification against case report forms. Independently, sponsors, institutions, and regulators conduct GCP audits, which differ from monitoring in scope and independence: audits are typically conducted by a party not involved in day-to-day trial conduct and assess systemic compliance rather than individual data points. Regulators including FDA (through its Bioresearch Monitoring, or BIMO, program) and EMA conduct their own GCP inspections of investigator sites, sponsors, and IRBs/IECs, with findings that can range from minor observations to trial data being disqualified or a site being barred from future regulated research.

GCP vs. GMP vs. GLP

GCP is one of several "GxP" quality standards that apply at different stages of drug and device development, and the three are easy to conflate:

• GCP governs how a clinical trial is conducted in human participants — informed consent, protocol adherence, investigator responsibilities, and data integrity.
• GMP (Good Manufacturing Practice; in the US, 21 CFR Parts 210/211 for drugs) governs how the investigational product itself is manufactured, tested, and released.
• GLP (Good Laboratory Practice; 21 CFR Part 58 in the US) governs the conduct and documentation of nonclinical (preclinical) safety studies, before human trials begin.

See our full comparison of GCP vs. GMP for a side-by-side breakdown of how the two standards apply to the same trial simultaneously but govern entirely different activities.

GCP vs. GCP certification — a common point of confusion

“Good Clinical Practice” is the regulatory standard itself; “GCP certification” is not a single government-issued credential built on top of it. No regulator issues a universal GCP license, and no single body owns the term “certified” in this context — what exists is a patchwork of course-completion certificates and professional credentialing exams from different training and credentialing organizations, all aimed at demonstrating competency in the same underlying ICH E6 principles. If you’re specifically evaluating training or credentialing options, see our dedicated guide to GCP certification.

Frequently asked questions

Is GCP a law or a guideline?

ICH E6 itself is a harmonized guideline, not a law. It becomes enforceable through the regulations of each implementing jurisdiction — for example, FDA regulations in the US and the EU Clinical Trials Regulation in the EU — which is why the practical compliance obligations can vary slightly by region even though the underlying GCP principles are shared.

Does GCP apply to all clinical research, or only trials of regulated products?

ICH E6 GCP specifically applies to interventional clinical trials of regulated medical products (drugs, biologics, and, via the related ISO 14155 standard, devices). Purely observational studies and non-regulated research are typically governed by separate human-subjects-protection frameworks, such as the Common Rule or institutional IRB policy, rather than by ICH GCP directly — though many institutions apply GCP-consistent practices more broadly as a matter of policy.

Who needs GCP training?

Anyone with a substantive role in the conduct, management, or oversight of a clinical trial — investigators, sub-investigators, study coordinators, and often sponsor and CRO staff — is generally expected to hold current GCP training. NIH-funded trials have an explicit training mandate (NOT-OD-16-148) with a three-year refresh cycle; many institutions and sponsors apply a similar cadence as internal policy even where no external mandate applies.

What’s the difference between GCP monitoring and a GCP audit?

Monitoring is an ongoing, sponsor-directed activity (often performed by a clinical research associate) that checks individual site data and procedures against the protocol as the trial progresses. An audit is a more independent, systemic review — conducted by a party not involved in day-to-day conduct — of whether the overall quality system met GCP requirements. Regulatory inspections are a third, distinct layer, conducted by the regulator itself.

Is ICH E6(R2) still relevant now that E6(R3) exists?

Yes, during the transition period. Implementation and effective dates differ by regulator and, in some cases, by which part of the guideline (principles vs. annexes) is involved, so trials that started under E6(R2) or that operate in a jurisdiction still phasing in E6(R3) may need to document compliance against the earlier revision for some time yet. Sponsors running multi-region trials should confirm the applicable revision with each relevant regulator rather than assuming a single global cutover date.

Canada is a worked example of exactly that staggered adoption: Health Canada is taking on E6(R3) at the same time as a proposed rewrite of its own trial regulations, so the applicable rule set depends on both the guideline revision and the domestic regulation in force. See Health Canada’s 2026 clinical trial overhaul.

See also: MHRA inspection — An MHRA GCP inspection is the UK Medicines and Healthcare products Regulatory Agency’s mechanism for verifying that clinical trials of medicines are conducted, recorded, and reported in accordance with Good Clinical Practice (GCP).