Standard Operating Procedure (SOP)

In a research context, a Standard Operating Procedure (SOP) is a written, controlled, step-by-step instruction for performing a specific recurring task the same way every time it is performed, regardless of who performs it -- distinct from a protocol, which specifies what a particular study will do, and from a policy, which states an institution's general position or requirement. A document counts as a research SOP only if it is: (1) task- or process-specific rather than study-specific (e.g. "how this laboratory calibrates a balance" or "how this site receives and stores investigational product," not "what Study ABC-123 will test"); (2) formally approved and version-controlled, with an identifiable author, approver, effective date, and version/revision history; (3) written in enough procedural detail that a trained staff member unfamiliar with that specific instance of the task could follow it and produce a consistent, auditable result; and (4) actually followed and enforced in practice, with deviations documented -- an unused or unenforced document is not functioning as an SOP even if it is labeled one.

ByCASRAI Editorial Board

· Last updated 18 Jul 2026

Worked examples

Is an instance

A GLP-compliant toxicology laboratory maintains a written SOP titled "Calibration and Standardization of Analytical Balances" specifying the calibration frequency, reference weights used, acceptance tolerances, the record log format, and who is authorized to perform and verify the calibration -- required under 21 CFR 58.81, which mandates written SOPs for equipment maintenance, testing, and calibration in every GLP-covered nonclinical laboratory study.

Is an instance

A clinical trial sponsor maintains an SOP titled "Adverse Event Reporting and Expedited Reporting Timelines" that applies across every trial the sponsor runs, describing how site staff document an adverse event, how it is assessed for seriousness and causality, and the internal timeline for escalating a Suspected Unexpected Serious Adverse Reaction (SUSAR) to pharmacovigilance -- the kind of sponsor-level quality-system SOP required under ICH E6(R2) Section 5.1.1, which obligates sponsors to implement quality assurance and quality control systems with written SOPs covering trial conduct and data handling.

Looks similar, but isn't

Not an instance

A study protocol describing the inclusion/exclusion criteria, dosing schedule, and endpoints for a single named clinical trial is not an SOP -- it defines what that one study will do, not a reusable, repeatable procedure for performing a task across studies. A protocol typically cites or incorporates the site's or sponsor's SOPs by reference (e.g. "adverse events will be reported per SOP-014") rather than restating the procedural detail itself.

Not an instance

A one-page lab policy stating that "all personnel must complete biosafety training annually" is a policy, not an SOP -- it states a requirement without specifying the step-by-step procedure (who schedules it, what the enrollment workflow is, how completion is documented and verified) that would make it operational and auditable.

A Standard Operating Procedure (SOP) is a written, controlled instruction that specifies exactly how a recurring task or process is to be performed, so that the result is consistent no matter who carries it out or when. In research settings — laboratories, clinical trial sites, research pharmacies, biorepositories, core facilities — SOPs are the operational backbone of a quality-management system: they are what turns a regulatory requirement or institutional policy into a repeatable, auditable, trainable set of actions.

This entry covers the general concept. For a fully worked, section-by-section example of one specific type of research SOP, see CASRAI’s Research Pharmacy SOP guide, which walks through investigational-product receipt, storage, accountability, blinding, dispensing, and destruction in a research pharmacy setting.

Why research units need SOPs: the regulatory basis

SOPs are not merely good practice — in regulated research, they are an explicit, enforceable requirement, and inspectors and auditors check for them directly.

Good Laboratory Practice (GLP), 21 CFR Part 58 (nonclinical/preclinical safety studies submitted to FDA) requires, at 21 CFR 58.81, that a testing facility have written SOPs covering routine processes that could affect study data integrity — animal care and receipt, test- and control-article handling and storage, equipment calibration and maintenance, data recording, and more. Deviations from an SOP during a study must be authorized by the study director and documented in the raw data; significant changes to an established SOP must be authorized in writing by management. See CASRAI’s Good Laboratory Practice (GLP) entry for the full scope of GLP.
Good Clinical Practice (GCP), ICH E6 imposes an equivalent obligation on the clinical trial side. ICH E6(R2) Section 5.1.1 requires the sponsor to implement and maintain a quality assurance and quality control system with written SOPs, so that trials are conducted and data generated, recorded, and reported in compliance with the protocol, GCP, and applicable regulatory requirements. ICH E6(R2) Section 1.38 defines “monitoring” itself partly by reference to SOPs: confirming a trial is conducted in accordance with the protocol, SOPs, GCP, and applicable regulations. Investigator sites are expected to be familiar with the sponsor’s SOPs as well as their own. See CASRAI’s ICH E6(R3) entry for the current core GCP guideline.
Good Manufacturing Practice (GMP) and other GxP frameworks impose parallel SOP requirements for their respective domains — see CASRAI’s Good Manufacturing Practice (GMP) guide.

Across all of these frameworks, the underlying logic is the same: a regulator or sponsor cannot verify that a study or trial was conducted correctly by trusting memory or informal habit alone. A written, approved, version-controlled SOP is the artifact that lets an auditor confirm the intended procedure existed, was followed, and that any deviation was documented and assessed.

SOP vs. protocol vs. policy: the distinction that trips people up

These three document types are frequently confused, but they answer different questions:

Policy answers what is required, at the institutional level, without procedural detail — e.g. “all research staff handling human biological specimens must complete biosafety training.”
SOP answers how a specific recurring task is performed, in step-by-step, trainable detail, independent of any single study — e.g. “how this laboratory logs, labels, and stores incoming specimens.” An SOP is reusable across studies and over time until formally revised.
Protocol answers what a specific study will do — objectives, design, population, interventions, endpoints, statistical plan — for one named study. A protocol commonly references the site’s or sponsor’s SOPs rather than restating their procedural detail (“specimens will be processed per SOP-022”). See CASRAI’s What Is a Research Protocol? guide for the full definition and structure of a protocol.

A useful test: if the document would still make sense with the study name removed and applies the same way to the next study that comes along, it is functioning as an SOP. If removing the study specifics leaves nothing meaningful, it is a protocol.

Typical SOP structure

Formats vary by institution and discipline, but a research SOP conforming to GLP/GCP quality-system expectations typically includes:

Header/control block: SOP title, unique identifier/number, version number, effective date, author, and approver(s) — usually a department head and a quality assurance (QA) reviewer, kept separate from the author.
Purpose and scope: what the SOP covers, and explicitly what it does not cover, plus the roles/personnel to whom it applies.
Definitions and abbreviations: terms used in the procedure that could otherwise be read ambiguously.
Responsibilities: who is authorized to perform each step, and who verifies or signs off on it.
Procedure: the step-by-step instructions themselves, detailed enough that a trained staff member unfamiliar with this specific instance of the task can follow them and produce a consistent result.
Records/documentation: what gets logged, on what form, and where it is retained.
References: the regulation, guideline, or institutional policy the SOP implements (e.g. the specific GLP or GCP citation).
Revision history: a table logging each version, its effective date, and a summary of what changed.

Version control and change management

An SOP that no one can prove was the current, approved version at the time a task was performed is a serious audit finding. Version control practices that regulated research units rely on include: a unique version number and effective date on every SOP; a formal review/approval workflow (author drafts, QA and/or a designated approver sign off) before a new version takes effect; a defined periodic review cycle (commonly annual, though the interval is institution-defined) even when no substantive change is needed, to confirm the SOP still reflects actual practice; retiring and archiving superseded versions rather than deleting them, since inspectors may need to confirm which version was in effect on a given date; and staff re-training and documented acknowledgment whenever a revision changes the procedure materially. Under GLP specifically, 21 CFR 58.81 requires that significant changes to an established SOP be authorized in writing by management, and that deviations during a study be documented in the raw data and authorized by the study director — an SOP is not treated as optional guidance once approved.

Frequently asked questions

Who writes a research SOP?

Typically the staff member or supervisor with direct operational responsibility for the task drafts it, with review and approval from a quality assurance function (where one exists) or a department/facility director. In a clinical trial, sponsor-level SOPs are usually written by the sponsor’s quality or clinical operations group; investigator sites may adapt or write site-specific SOPs that operate alongside the sponsor’s.

Does every research task need its own SOP?

No — GLP and GCP require SOPs for processes that could affect data integrity, safety, or compliance, not for every conceivable action. The judgment of which processes warrant a formal SOP is itself typically documented in a facility’s or sponsor’s quality system.

How is an SOP different from a work instruction or job aid?

Terminology varies by institution, but many quality systems distinguish an SOP (the formally controlled, approved procedure) from a less formal work instruction or quick-reference job aid that summarizes the SOP for day-to-day use without itself being the controlled document of record.

Do SOPs apply outside regulated GLP/GCP research?

Yes — many research units maintain SOPs as good practice even where no specific regulation mandates them, because the underlying benefit (consistency, trainability, auditability) applies to any research process with room for procedural drift, including core facilities, biorepositories, and data management workflows.