GCP record retention: how long to keep clinical trial records and the TMF

At a glance

• The FDA requires an investigator to keep investigational drug study records for 2 years after a marketing application is approved for the indication studied, or, if no application is filed or it is not approved, for 2 years after the investigation is discontinued and FDA is notified (21 CFR 312.62(c)).
• The sponsor’s FDA clock is also 2 years, but it runs from approval of the drug, or from the end of investigational shipments plus notice to FDA (21 CFR 312.57(c)).
• Under EU Regulation 536/2014, sponsor and investigator archive the content of the trial master file for at least 25 years after the end of the trial (Article 58).
• ICH E6(R3) sets no number. The investigator keeps essential records for the period applicable regulations require, or until the sponsor says they are no longer needed, whichever is the longest (§2.12.12).
• Retention is a condition as well as a period: records must stay complete, readable, readily available and directly accessible to regulators, monitors and auditors (ICH E6(R3) C.2.6).

There is no single GCP record retention period. ICH E6(R3) leaves the number to the law that governs your trial, so the answer to “how long do we keep clinical trial records” is 2 years from a defined trigger under FDA’s Part 312, at least 25 years from the end of the trial under the EU Clinical Trials Regulation, and longer still for some EU marketing and advanced therapy cases. The table below gives each period with its source, then the sections explain the triggers people miscount and what archiving the trial master file (TMF) requires.

Retention periods by rule

Rule	Who it binds	Minimum period	Clock starts
21 CFR 312.62(c)	Investigator	2 years	Approval of a marketing application for the indication investigated. If none is filed or it is not approved: discontinuation of the investigation and notice to FDA.
21 CFR 312.57(c)	Sponsor	2 years	Approval of a marketing application for the drug. If not approved: end of shipment and delivery of the drug for investigational use, and notice to FDA.
ICH E6(R3) §2.12.12	Investigator / institution	No fixed number	The period applicable regulations require, or until the sponsor says records are no longer needed, whichever is the longest.
ICH E6(R3) §3.16.3(a)	Sponsor	No fixed number	Per the applicable regulatory requirements.
EU Regulation 536/2014, Article 58	Sponsor and investigator	At least 25 years	End of the clinical trial. Subjects’ medical files follow national law.
EMA TMF guideline §6.3 (Directive 2005/28/EC)	Trials run under Directive 2001/20/EC	At least 5 years	Completion of the trial.
EMA TMF guideline §6.3 (Directive 2003/63/EC)	Trials whose data support an EU marketing authorisation	At least 15 years, or 2 years after the last EU marketing authorisation, or 2 years after formal discontinuation of clinical development, whichever is longest	Completion or discontinuation of the trial. Some core documents are kept as long as the product is authorised.
EMA TMF guideline §6.3	Documents for full traceability of an advanced therapy (ATIMP)	30 years, or longer if the trial authorisation requires	Expiry date of the product.

Two rows trip people up. A 15-year figure is not an FDA rule: it is the EU marketing authorisation period from Directive 2003/63/EC, as the EMA guideline describes it, while Part 312 states 2 years. And the EU 25-year period in Article 58 is a floor: it applies “unless other Union law requires archiving for a longer period”.

The FDA answer: 2 years, counted from approval or discontinuation

The FDA requires retention of investigational drug study records for 2 years following the date a marketing application is approved for the drug for the indication being investigated. If no application is to be filed, or the application is not approved for that indication, the records are kept until 2 years after the investigation is discontinued and FDA is notified (21 CFR 312.62(c)).

If you are answering this as an exam question, the option that matches the regulation is “at least two years after the investigational drug’s approval by the FDA”. The other usual options, “one year after the last participant completes”, “three years after the study ended at the site” and “indefinitely”, are not the Part 312 rule.

Three details change the date you write in the retention schedule:

• The clock does not start at the end of the trial. For the investigator, it starts at approval for the specific indication studied. A site that closed out years before approval keeps its records until 2 years after that approval.
• The fallback clock needs a notification. Where no application is filed, the 2 years run from discontinuation and notice to FDA. An investigation that stopped without that notice has not started its fallback clock.
• The sponsor’s trigger differs from the investigator’s. The sponsor counts from approval of a marketing application for the drug or, if one is not approved, from the end of shipment and delivery of the drug for investigational use plus notice to FDA (21 CFR 312.57(c)).

The investigator’s retention duty covers records required under Part 312. Those include drug disposition records with dates, quantities and use by subjects (21 CFR 312.62(a)), and case histories, which include case report forms, signed and dated consent forms and medical records such as progress notes and hospital charts (21 CFR 312.62(b)). Throughout the period, an investigator must let an authorised FDA officer access, copy and verify those records at reasonable times (21 CFR 312.68), and a sponsor must do the same for records relating to the investigation (21 CFR 312.58(a)).

ICH E6(R3) sets no number

E6(R3) is the guideline people most often search for a period in, and it names none. The investigator or institution retains essential records for the required retention period in accordance with applicable regulatory requirements, or until the sponsor informs them the records are no longer needed, whichever is the longest (ICH E6(R3) §2.12.12). The sponsor is asked to obtain the investigator’s agreement to exactly that (§3.6.3(c)).

For the sponsor, E6(R3) asks three things about retention:

• Retain the sponsor-specific essential records in accordance with the applicable regulatory requirements (§3.16.3(a)).
• Tell investigators, institutions and, where appropriate, service providers in writing what the retention requirements are, and notify them in writing when the records are no longer needed (§3.16.3(b)).
• Make sure the investigator has access to the data needed for retention (§3.16.1(m)), and protect trial data from unauthorised alteration and from inappropriate destruction or accidental loss (§3.16.1(v)).

The practical effect is that a site’s destruction date comes from two places: the law, and the sponsor’s written notice. A site that destroys records because “the 2 years are up” without hearing from the sponsor is working from half the rule.

The EU answer: at least 25 years after the end of the trial

Under Article 58 of Regulation 536/2014, unless other Union law requires a longer period, the sponsor and the investigator archive the content of the clinical trial master file for at least 25 years after the end of the clinical trial. The medical files of subjects are archived in accordance with national law instead (Article 58).

The trigger is the end of the trial, and the Regulation defines what counts for a trial that stops early. On early termination, the date of the early termination is deemed to be the end of the trial (Article 37(7)). A temporarily halted trial that does not resume within two years ends at the expiry of that period or at the sponsor’s decision not to resume, whichever is earlier (Article 37(7)).

The EMA TMF guideline asks the sponsor to work out which requirements apply from the trial’s start and end dates and whether its data support, or are intended to support, a marketing authorisation (EMA TMF guideline §6.3). That is why the table carries three EU rows: older trials under Directive 2001/20/EC, trials supporting a marketing authorisation, and ATIMP traceability. The same section applies these retention times to documents held by CROs and other agents of the sponsor (§6.3).

The US and EU periods do not line up

Part 312 and the EU Regulation give different numbers measured from different events. Part 312 sets 2 years from approval, or from discontinuation and notice to FDA (21 CFR 312.62(c), 312.57(c)). Article 58 sets at least 25 years from the end of the trial. Neither text refers to the other or says how to apply the two together for a trial that runs in both places.

ICH E6(R3) does not settle it either. Its “whichever is the longest” clause compares the regulatory requirement with the sponsor’s notice (§2.12.12). It does not say how to combine two jurisdictions. A trial with US and EU sites therefore has two retention rules applying to records that may sit in one archive, and the sponsor’s decision on which records fall under which rule is what explains every destruction date. Write that decision down with the provisions it relies on, because the people who made it will have moved on long before the destruction date.

What “retained” means beyond the number of years

Keeping records for the full period while letting them become unreadable meets neither E6(R3) nor the EU Regulation. E6(R3) states the condition: sponsor and investigator retain essential records so that they remain complete, readable and readily available, and are directly accessible upon request by regulatory authorities, monitors and auditors. Alterations must be traceable (ICH E6(R3) C.2.6). The investigator also takes measures to ensure availability, accessibility and readability, and to prevent unauthorised access and accidental or premature destruction (§2.12.12).

The EU Regulation adds the same expectations in its own words. The archived TMF content must be readily available and accessible to the competent authorities on request, the archive media must keep the content complete and legible for the whole period, and any alteration must be traceable (Article 58).

Each condition fails in its own way. A file missing the amendment that superseded a document is incomplete. Thermal paper printouts fade. Boxes in deep offsite storage with a multi-week recall are hard to call readily available. An archive whose only password left with a former employee is not accessible to anyone.

TMF archiving: named owners, restricted access, a reviewed file

An archive is defined by who controls it. Article 58 requires the sponsor to appoint individuals within its organisation to be responsible for archives, and restricts access to those individuals. The EMA guideline adds that their appointment and training should be documented, withdrawal of documents from the archive should be under their control, and an archive index or log should record every TMF entered (EMA TMF guideline §6.1).

Before anything goes into the archive, the investigator or institution and the sponsor should review that their filed TMF documentation is complete. Access to the archive should be restricted, and archived electronic documents protected from unauthorised changes (EMA TMF guideline §6). If you need a refresher on what the TMF contains in the first place, see what the trial master file is, and use the TMF checklist for the completeness review itself.

The investigator’s side has its own rules. The investigator or institution tells the sponsor how its essential documents are stored, keeps ultimate responsibility for them even when they sit in an external archive, and keeps a list of all trials and their archiving arrangements (EMA TMF guideline §6.2). If the investigator can no longer be responsible, for example through relocation, retirement or closure of the institution, the sponsor should be notified in writing and told who takes over (§6.2). E6(R3) puts the same duty in shorter form: keep the sponsor informed of the name of the person responsible for the essential records during the retention period, for example when the site closes or an investigator leaves (§2.12.13). The split between the sponsor file and the investigator site file is covered in TMF vs ISF.

External archives are allowed. The sponsor or investigator should assess the facility before use, keep quality assurance going afterwards, and have a formal agreement that ideally covers the sponsor or the archive going out of business (EMA TMF guideline §6). When ownership of the TMF content transfers, the transfer must be documented and the new owner takes on the Article 58 responsibilities (Article 58).

eTMF retention: the archive has to outlive the software

A 25-year period can outlast both the eTMF software and the vendor contract, so eTMF retention is a question of keeping content readable after systems change. The EMA guideline states that access to documents and data must be maintained for the entire archiving period. That may mean keeping the original hardware and software, emulating the old software, or migrating the data to a new format, and the organisation should address this in written procedures (EMA TMF guideline §6). The audit trail is archived with the eTMF and its dynamic character preserved where applicable (§6).

Migration is where content gets lost. Any transfer or migration must be validated so the migrated data hold the same information as the original, including metadata (EMA TMF guideline §6). E6(R3) asks for the same when electronic data move between computerised systems, with documented, traceable transfers and reconciliation to avoid data loss (§4.2.5), and for archived trial data and metadata to stay retrievable, readable and protected from unauthorised changes for the whole retention period (§4.2.7).

Questions to settle while you still have the vendor relationship:

• In what format does content leave the system, and can it be read without the vendor’s software?
• Does an export carry the audit trail and metadata, or only the documents?
• What does the contract say if the vendor is acquired or stops trading?
• Who in your organisation is named as responsible for the archive, and who replaces them?

For choosing a system with these answers in mind, see choosing an eTMF system that survives a GCP inspection.

Scanning paper into an eTMF can shorten how long you keep the paper. Creating certified copies that meet the guideline’s requirements could enable earlier destruction of the originals, but essential documents must not be destroyed before the end of the required retention period (EMA TMF guideline §5.5). The digitisation process must be validated so that no information is lost or altered (§5.4).

When records can be destroyed

Under E6(R3), trial data and metadata may be permanently destroyed when no longer required as determined by applicable regulatory requirements (§4.2.8). For the site, the release comes from the sponsor: it is the sponsor’s responsibility to inform the investigator or institution in writing when trial documents no longer need to be retained (EMA TMF guideline §6.3), and E6(R3) asks the same written notice of the sponsor (§3.16.3(b)).

A workable retention schedule records, for each trial: the rules that apply and why, the trigger event for each, the date that event happened (or “pending”, for an approval that has not come), the resulting earliest destruction date, and the written sponsor notice when it arrives. Close-out is when most of these fields get filled in, so the clinical trial close-out checklist is the natural place to start the record.

Sources

• ICH E6(R3) Good Clinical Practice, version R3
• 21 CFR Part 312, Investigational New Drug Application (current as of 4/14/2026)
• Regulation (EU) No 536/2014 on clinical trials on medicinal products for human use
• EMA Guideline on the content, management and archiving of the clinical trial master file (paper and/or electronic), EMA/INS/GCP/856758/2018