Guides · Everyone

How long to keep veterinary study records

Two years after the study ends or the drug is approved under 21 CFR 511.1, five years for adverse event reports under 514.80, and what must survive.

Published November 6, 2026 · sources checked November 6, 2026. Signalment is a data system; the protocol, the ethics review, the consent and any filing stay with the sponsor and the investigators.

For a study of an unapproved animal drug run under an investigational exemption, the clock is set by 21 CFR 511.1: the investigator keeps complete records of the investigation for two years after the investigation ends or a new animal drug application is approved, and the sponsor keeps every report it received from investigators for the same period. The rule names both events and gives no order of priority, so the careful reading is the later of the two, and an approval can come years after the last visit. Adverse event records for an approved product are kept five years after the report is submitted (21 CFR 514.80(e)). An observational study with nothing assigned has no federal clock at all; its protocol is the rule. Whatever the period, the audit trail for an electronic record must last at least as long as the record (21 CFR 11.10(e)).

Which rule sets the clock

RecordRulePeriod
The investigator's records of the investigation, including receipt and disposition of each shipment of the drug21 CFR 511.1(b)(7)(ii)2 years after the investigation ends or a new animal drug application is approved
The reports the sponsor receives from investigators21 CFR 511.1(b)(8)(i)The same period, and open to FDA inspection
The sponsor's record of each shipment, with the investigator's name and address21 CFR 511.1(b)(3)2 years after the shipment
Records and reports of adverse drug experiences for an approved product21 CFR 514.80(e)5 years after the date of submission
A veterinarian's records of extra-label use of an approved drug21 CFR 530.5(b)2 years, or longer where federal or state law requires it
The audit trail of an electronic record kept under an FDA requirement21 CFR 11.10(e)At least as long as the record it documents
All study documentation under good clinical practiceVICH GL9, 8.1.2 and 8.4.3The period required by the regulatory authority the study is submitted to

Two rules in that table set no number. VICH GL9 section 8.1.2 says all study documentation should be retained "for the period of time required by relevant regulatory authorities," and 8.4.3 ties that period to the authorities "to which the study may be or has been submitted." For a study meant for FDA, the number comes from 511.1. Part 11 is the same: section 11.10(c) asks for "protection of records to enable their accurate and ready retrieval throughout the records retention period," and the period comes from the predicate rule. FDA's October 2024 questions-and-answers guidance on electronic systems, which the Center for Veterinary Medicine co-issued, names 511.1(b)(7)(ii) and 511.1(b)(8)(i) as those predicate rules in question 5.

The clinic's own medical record runs on a different clock. State veterinary practice acts set a period for the practice's patient record, counted from the animal's last visit, and the period differs by state. The practice's licensing board publishes it. That record belongs to the practice; the study record is a separate thing with its own rule.

How the design changes the answer

What has to survive

GL9 section 8.2 lists what study documentation is. The protocol with every amendment and every deviation. The raw data: animal records (inclusion, exclusion and removal, the owner's informed consent, treatment assignment, every recorded observation, case report forms, adverse events, final disposal), product records (ordering, receipt, inventory, administration, return, disposal), the monitor's and investigator's contact records, and facility and equipment records. The reports: safety reports, the final study report and the statistical, analytical and laboratory reports. The standard operating procedures and reference materials. Section 8.4.1 asks that all of it be stored protected from "deterioration, destruction, tampering or vandalism," and 8.4.2 that the final study report state where it is.

For electronic records, the 2024 guidance adds the parts people forget:

The 2003 Part 11 scope guidance, which CVM also issued, says FDA does not object to archiving required records to paper or to a standard file format such as PDF or XML, and to deleting the electronic version afterwards, as long as the predicate rule is satisfied and the content and meaning of the records are preserved. An archive that drops the audit trail or the metadata does not preserve the meaning of the record.

Steps before a study closes

  1. Write the retention period and the storage location into the protocol. GL9 section 6.3.18 asks the protocol to specify procedures for "recording, processing, handling, and retaining raw data and other study documentation."
  2. Decide the date the clock starts. For an INAD study, record the date of termination, and note that approval of the application may restart the count.
  3. Take the complete export: one file per form, the data dictionary, the consents, the signatures, the queries and the full audit trail, with the fingerprints a reviewer can recompute. Record the date and time of the export so it stands as a certified copy.
  4. Keep the product records with the study. Receipt and disposition of every shipment are named in 511.1(b)(7)(ii).
  5. State in the final study report where the documentation and any authenticated copy are held (GL9 8.4.2).
  6. Check the vendor agreement. Question 17 of the 2024 guidance says the agreement with an IT service provider should include a plan that ensures the sponsor has access to the data throughout the regulatory retention period.
  7. Put the end date in a calendar that outlives the study team.

Whose job it is

The protocol, the ethics review, the consent content, any filing with FDA and any report of an adverse event or a significant hazard stay with the sponsor and the investigator. GL9 section 4.2.14 puts the duty to maintain study documentation for the required period on the sponsor, and 3.2.31 puts secure storage of the investigator's copy on the investigator. Signalment is a data system. It keeps the records attributable and versioned, exports them as CSV per form, a data dictionary and a JSON bundle with the full audit trail, and closes a study with a lock that carries a fingerprint, so a monitor can verify the archived copy offline. It does not yet keep a drug accountability ledger, so the shipment receipt and disposition record that 511.1 names is kept alongside the export, and it does not yet produce SAS or SEND files. How the lock and the export work is in How to lock a study database; the sponsor page at signalment.app/sponsors shows the export checked on a laptop.

Sources

Your next study

See it with your own study.

A 20-minute demo: an example study across four clinics, then how yours would be set up. Or start from a template in a free workspace.