The FDA IND Application as a Three-Pillar Safety Dossier: 21 CFR 312.23 Contents, the 30-Day Clock, and the Sponsor-Investigator Traps That Trigger a Clinical Hold

At a glance

• An IND is not a Form 1571 checklist. It is a safety argument built from three pillars (CMC, pharmacology/toxicology, and the clinical protocol) that FDA reads to decide whether your first human subjects face an unreasonable risk.
• 21 CFR 312.23 sets the content and format; each section answers a specific safety question the reviewer is asking, and a gap in any one of them is grounds for a clinical hold under 312.42(b).
• The IND goes into effect 30 days after FDA receives it unless FDA imposes a clinical hold or notifies you earlier that you may proceed. There is no approval letter for a default start.
• There is no IND user fee. PDUFA fees attach to marketing applications, not investigational ones.
• Form FDA-1571 is the sponsor’s cover sheet; Form FDA-1572 is the investigator’s signed statement. They are different documents with different signers, and conflating them is a common avoidable error.
• Sponsor-investigators (the academic IND path) carry the duties of both sponsor and investigator, and most preventable hold letters trace to thin pharm/tox bridging, vague protocol stopping rules, or missing CMC stability data.

A first-time sponsor-investigator tends to treat the IND as paperwork: fill in the cover sheet, attach the protocol, mail it in, wait. That framing is how holds happen. FDA does not review an IND to confirm you filled out the forms. It reviews it to answer one question for each section of your submission: is there enough information here to conclude that the human subjects you are about to dose are not exposed to an unreasonable and significant risk? This article maps the anatomy of the IND to that question, section by section, so you can judge whether your dossier is review-ready before the clock starts.

When you actually need an IND (and the 312.2 exemptions)

The default rule is broad. Under 21 CFR 312.2(a), the part applies to all clinical investigations of products subject to section 505 of the Federal Food, Drug, and Cosmetic Act or the licensing provisions of the Public Health Service Act. In plain terms: if you are studying a drug or biologic in humans, assume you need an IND until you have confirmed an exemption applies.

The narrow off-ramp is 312.2(b). The clinical investigation of a drug product that is lawfully marketed in the United States is exempt only if all of several conditions hold together: the study is not intended to support a new indication or other significant labeling change, it does not involve a route, dose, or population that significantly increases risk, and it is conducted under the institutional-review (part 56) and informed-consent (part 50) requirements. The trap here is reading the list as “any one of these.” It is conjunctive. A study of a marketed drug at a higher dose, or in a sicker population, falls outside the exemption even if every other condition is met. When in doubt, the conservative read is that you need an IND.

The three pillars of every IND

It helps to stop thinking of the IND as a stack of forms and start thinking of it as three evidentiary pillars that together support one conclusion: this drug is reasonably safe to put into the specific humans described in the protocol.

1. Chemistry, manufacturing, and controls (CMC). What is the drug, how is it made, and how do you know each batch is what you say it is? 21 CFR 312.23(a)(7) requires information on the drug substance and product, including the acceptable limits and analytical methods used to assure identity, strength, quality, and purity. The reviewer’s question: could a quality or stability failure harm a subject?
2. Pharmacology and toxicology. What did the animal and in vitro data show, and do they support the proposed human dose and duration? 21 CFR 312.23(a)(8) requires pharmacology and toxicology information adequate to conclude that the drug is reasonably safe to conduct the proposed clinical investigations. The reviewer’s question: does the nonclinical package bridge to first-in-human exposure?
3. The clinical protocol. Who gets dosed, how much, monitored how, and stopped when? The protocol is where the other two pillars meet the patient. The reviewer’s question: is the plan to protect subjects credible and specific?

ICH E8(R1) §3.1 frames this from the design side: quality by design means building quality into the study protocol and processes proactively rather than inspecting for it later. The IND is the first place a reviewer sees whether you did that.

The 312.23 contents, mapped to the safety question FDA is asking

21 CFR 312.23(a) lists the IND contents in a specific order. The format set out in 312.23 should be followed routinely by sponsors in the interest of fostering efficient review. Read each section not as a box but as an answer to a reviewer’s question.

312.23(a) section	What it contains	The safety question the reviewer is asking
(a)(1) Cover sheet (Form FDA-1571)	Sponsor identity, phase, commitments (IRB review, not starting before the IND is in effect)	Who is responsible, and have they committed to the basic guardrails?
(a)(2) Table of contents	Navigation	Can a reviewer find the safety-critical data fast?
(a)(3) Introductory statement and general investigational plan	The drug, prior human experience, the plan for the coming year	What is already known, and where is this going?
(a)(5) Investigator brochure	Consolidated safety and pharmacology summary for investigators	Do the people dosing subjects have an accurate risk picture?
(a)(6) Protocols	Study design, population, dosing, monitoring, stopping rules	Is the plan to protect subjects specific and credible?
(a)(7) CMC information	Composition, manufacture, controls, stability	Could a manufacturing or quality failure harm a subject?
(a)(8) Pharmacology and toxicology	Nonclinical data supporting the proposed human exposure	Does the animal package bridge to the first human dose?
(a)(9) Previous human experience	Summary of prior human use, if any	Has anything already gone wrong in people?

Per 21 CFR 312.23(c), the central focus of the initial IND submission should be the general investigational plan and the protocols for specific human studies. That is a direct statement of where the reviewer’s attention goes first. A polished cover sheet over a vague protocol is a misallocation of effort.

The protocol pillar carries an extra weight from the GCP and study-design guidelines. ICH E6(R3) Principle 4 requires that clinical trials be scientifically sound and based on adequate and current scientific knowledge. ICH E6(R3) Principle 8 requires that trials be described in a clear, concise, scientifically sound, and operationally feasible protocol. ICH E8(R1) §3.2 adds that factors critical to the quality of the study (the attributes fundamental to protecting participants and to the reliability of results) should be identified prospectively. These three say, in different words, the same thing: the protocol in your IND has to demonstrate design quality, not just describe procedures.

Forms decoded: 1571 vs 1572 vs 3674

The single most common clerical confusion in a first IND is treating the cover sheet and the investigator statement as interchangeable. They are not.

Form	Purpose	Who signs	When filed
FDA-1571	The IND cover sheet. Identifies the sponsor and carries the sponsor’s core commitments.	The sponsor (or sponsor-investigator)	As item (a)(1) of the IND, with the initial submission
FDA-1572	The investigator’s signed statement of qualifications and commitments to conduct the study per protocol and the regulations.	Each participating clinical investigator	Obtained by the sponsor before the investigator begins participation
FDA-3674	Certification regarding the clinicaltrials.gov registration requirements (registry/results-database obligations).	Sponsor / responsible party	With applicable submissions

Two of these are anchored in the corpus. 21 CFR 312.23(a)(1) requires the cover sheet to be Form FDA-1571 and lists the commitments it carries, including a commitment not to begin clinical investigations until the IND is in effect and a commitment that an IRB complying with part 56 will review and approve each study. Separately, 21 CFR 312.53(c) requires the sponsor, before permitting an investigator to begin participation, to obtain a signed investigator statement (Form FDA-1572) in which the investigator commits to conduct the study in accordance with the current protocol and to comply with the obligations of clinical investigators in part 312. The 1571 is the sponsor speaking; the 1572 is each investigator speaking. (Form FDA-3674 and the clinicaltrials.gov certification sit outside the in-scope corpus here; treat the row above as orientation and confirm current registry obligations against FDA’s own current instructions before you rely on it.)

For sponsor-investigators, this distinction collapses onto one person but does not disappear: you sign the 1571 as the sponsor and the 1572 as the investigator, and you are accountable under both sets of duties.

The 30-day clock: may proceed, silent default, and clinical hold

This is the part most “complete guide” pages bury in a footnote, and it is the part that decides whether you can dose your first subject.

Under 21 CFR 312.42(b), an IND goes into effect either thirty days after FDA receives the IND, unless FDA notifies the sponsor that the investigations are subject to a clinical hold, or on earlier notification by FDA that the clinical investigations may begin. There are therefore three outcomes, and only one of them produces a letter telling you to start.

Outcome	What happened	What you may do
Early “may proceed”	FDA affirmatively notifies you before day 30 that the investigations may begin	Begin dosing on that notification
Silent default	Day 30 passes with no clinical hold and no affirmative notice	The IND is in effect; you may begin. Do not expect an approval letter
Clinical hold	FDA issues a clinical hold order under 312.42	You may not begin. When a study is on clinical hold, subjects may not be given the investigational drug

A few practitioner points the corpus makes explicit. First, an investigator may not administer the investigational drug to human subjects until the IND goes into effect under 312.42(b). The 30-day clock is not advisory; it is a gate. Second, under 21 CFR 312.42(c), FDA communications with a sponsor that are not accompanied by a clinical hold order are solely advisory and do not require any modification to your plan. A reviewer’s phone call asking a clarifying question is not a hold. Third, no more than 30 days after imposing a clinical hold, the FDA division director will provide the sponsor a written explanation of the basis for the hold, so a hold is never a black box.

The grounds for a hold are where the safety argument and the paperwork meet. Under 21 CFR 312.42(b)(1), FDA may place a Phase 1 study on clinical hold if human subjects are or would be exposed to an unreasonable and significant risk; if the named clinical investigators are not qualified by scientific training and experience; if the investigator brochure is misleading, erroneous, or materially incomplete; or if the IND does not contain sufficient information required under 312.23 to assess the risks to subjects. For Phase 2 or 3, 21 CFR 312.42(b)(2) adds a hold ground where the protocol is clearly deficient in design to meet its stated objectives. Notice that two of these grounds (investigator qualification and insufficient 312.23 information) are not about the science of the molecule at all. They are about whether your dossier is complete and your investigators are credible.

Fees and the public record

Two honest answers that the search queries demand.

There is no IND user fee. The corpus governing INDs (21 CFR Part 312) imposes no application fee, and PDUFA user fees attach to marketing applications (NDAs and BLAs), not to investigational applications. If you have budgeted an “IND fee,” reallocate it. This is non-regulatory clarification rather than a quotable requirement, but it is the correct answer.

There is no fully public IND dossier database. INDs are not published as open records the way some marketing-application reviews are. What is public is downstream: trial registration and summary results on clinicaltrials.gov, and whatever FDA itself publishes. If you are looking for an “IND database” to model your submission on, you will not find one; rely on FDA’s current guidance and the regulation text instead.

Sponsor-investigators: the academic path and where it fails

A sponsor-investigator is defined in 21 CFR 312.3 as an individual who both initiates and conducts an investigation, and under whose immediate direction the investigational drug is administered or dispensed. That definition is the whole problem in one sentence: you wear both hats, so you owe both sets of duties.

The regulation anticipates the academic path. 21 CFR 312.23(d) says a sponsor-investigator who uses, as a research tool, a drug already subject to a manufacturer’s IND or marketing application should follow the same general format but may, if authorized by the manufacturer, refer to the manufacturer’s IND for the technical (CMC and pharm/tox) information. If there is no manufacturer’s IND to reference, the sponsor-investigator is ordinarily required to submit all the supporting technical information themselves, unless it can be referenced from the scientific literature. That single fork is where most academic INDs succeed or stall: a right of reference turns the CMC and pharm/tox pillars into a letter; its absence turns them into a full data package you must assemble.

On the sponsor side of the hat, 21 CFR 312.50 makes sponsors responsible for selecting qualified investigators, ensuring proper monitoring, and ensuring the investigation is conducted per the general investigational plan and protocols. ICH E6(R3) §2.1.1 reinforces the investigator side: the investigator should be qualified by education, training, and experience and should provide evidence of those qualifications. As a sponsor-investigator you are both the selector and the selected, which means the qualification evidence FDA looks for is your own CV and your own resources.

Where sponsor-investigators get it wrong

• Thin pharm/tox bridging. Submitting nonclinical data that does not clearly support the specific proposed human dose and duration. 21 CFR 312.23(a)(8) asks for enough to conclude the drug is reasonably safe at the proposed exposure, and “reasonably safe” is the reviewer’s bar, not a literature citation count.
• Vague protocol stopping rules. A protocol that describes dosing in detail but is hand-wavy about what triggers pausing or stopping. Under 312.42(b)(2), a protocol clearly deficient in design is a hold ground; ICH E6(R3) Principle 8 expects a clear, scientifically sound protocol.
• Missing CMC stability data. Treating the drug as a given. 21 CFR 312.23(a)(7) wants the analytical methods and limits that assure identity, strength, quality, and purity, and stability is part of “you know it is still the drug when you dose it.”
• Assuming a referenced IND covers everything. A manufacturer’s right of reference under 312.23(d) covers the technical information it authorizes, not your protocol, your investigator qualifications, or your IRB arrangements. Those remain yours.
• Treating the 30-day default as an approval. Waiting for a “may proceed” letter that, on a silent default, never comes, or worse, dosing before day 30. The investigator may not administer the drug until the IND is in effect under 312.42(b).

A note on how these guidelines fit together rather than fight: 21 CFR Part 312 sets the binding U.S. content and timing rules; ICH E6(R3) and ICH E8(R1) describe the GCP and study-design quality the protocol pillar must demonstrate. They point the same direction here. The regulation tells you what to file and when you may proceed; the ICH guidelines tell you what “scientifically sound protocol” means in practice. No tension between them surfaced for the claims in this article.

For the adjacent mechanics, see the sibling guides on the clinical trial sponsor’s responsibilities, the FDA Form 1572 in depth, principal-investigator responsibilities under GCP, the clinical trial protocol, and clinical trial safety reporting. This article deliberately stops at content and decision logic; the step-by-step eCTD/gateway submission lives with the IND-process guide.

Software and well-built processes can make assembling and tracking an IND far less error-prone, but they do not make a submission complete or a sponsor compliant. The regulations describe what FDA requires; meeting those requirements, and standing behind the safety argument, remains the sponsor’s responsibility.

Sources

• 21 CFR Part 312, Investigational New Drug Application — FDA, version 2026-04
• ICH E6(R3) Good Clinical Practice — ICH, version R3 (2025)
• ICH E8(R1) General Considerations for Clinical Studies — ICH, version R1 (2021)