“Clinical trial phases,” “clinical trial stages,” and “clinical research phase” all point to the same classification system: the sequence FDA uses to describe how far a drug or biologic has progressed through human testing. The phase a trial is in is not a formality — it determines the regulatory filing required to start it, the number and type of participants enrolled, how intensively the trial is monitored, and how heavy the safety-reporting burden is on the sponsor and site. For a research administrator, phase classification is one of the first facts that shapes staffing plans, budget, and monitoring visit frequency for a given protocol, well before the science itself is the deciding factor.
This guide walks through FDA’s phase framework — Early Phase 1 (Phase 0) through Phase 4 — the regulatory milestones that separate each phase (the Investigational New Drug application, or IND, before human testing begins; the New Drug Application or Biologics License Application, NDA/BLA, after Phase 3), and what phase classification actually changes about how a trial is administered.
The five-phase framework FDA and ClinicalTrials.gov use
The legal basis for the phase system is 21 CFR 312.21, “Phases of an investigation,” which defines Phase 1, Phase 2, and Phase 3 for drugs and biologics regulated under an Investigational New Drug (IND) application. The regulation is explicit that these phase descriptions are not rigid categories — “the clinical trial process…generally consists of four phases,” and “the phases of an investigation…are not necessarily conducted in strict sequence. Overlap…frequently occurs.” Phase 4 (postmarketing) sits outside 312.21 and is instead addressed by FDA’s postmarketing requirements and commitments framework.
ClinicalTrials.gov’s protocol registration system uses five phase designations: Early Phase 1 (formerly labeled Phase 0), Phase 1, Phase 2, Phase 3, and Phase 4. This is the taxonomy referenced throughout this guide, and it is the same taxonomy behind the “clinical trial stages” and “clinical research phase” phrasings — they describe the identical classification, not a separate one.
| Phase | Primary purpose | Typical participant count | Typical duration | Governing IND? |
|---|---|---|---|---|
| Early Phase 1 (Phase 0) | Exploratory pharmacokinetics/pharmacodynamics at sub-therapeutic doses | Roughly 10–15 | Days to a few weeks | Exploratory IND (reduced preclinical package) |
| Phase 1 | First-in-human safety, tolerability, dosing, metabolism | Roughly 20–80 | Several months to about a year | Standard IND |
| Phase 2 | Preliminary efficacy signal, common short-term risks, dose-finding | Roughly 100–300 | About 1–2 years | Standard IND (may be same or expanded IND) |
| Phase 3 | Confirmatory (pivotal) efficacy and safety in a larger, more representative population | Roughly 300 to several thousand | About 1–4 years | Standard IND |
| Phase 4 | Post-approval surveillance and additional studies | Varies; can exceed pre-approval enrollment | Ongoing after approval, sometimes years | Not IND-governed; conducted under the approved NDA/BLA |
The participant ranges for Phase 1–3 come directly from 21 CFR 312.21. The duration ranges are typical/illustrative, not a fixed regulatory requirement — actual timelines vary substantially by therapeutic area, endpoint, and enrollment rate; treat them as planning ranges, not commitments.
Early Phase 1 (Phase 0): exploratory IND studies
Early Phase 1 studies, formally Exploratory IND studies (a category FDA created in 2006), use very limited human exposure — often microdoses, defined as roughly 1/100th of the pharmacologically active dose estimated from animal data — with no therapeutic or diagnostic intent. The goal is narrowly mechanistic: does the compound behave in humans the way preclinical data predicted, before committing to a full Phase 1 program. Because exposure is sub-therapeutic and brief, FDA allows a reduced preclinical toxicology package relative to a standard IND, which is the main reason sponsors use this pathway: it screens out unpromising candidates earlier and cheaper than a full Phase 1 study would.
Administratively, Phase 0 trials are small enough that a single coordinator can often manage the entire study, but the science requires specialized bioanalytical and imaging capability (many Phase 0 designs use PET microdosing or similarly sensitive assays) that most general clinical sites do not have in-house — staffing is a capability question more than a headcount question.
Phase 1: first-in-human safety
Phase 1 is where a drug is introduced into humans for the first time (excluding any Phase 0 work). Per 312.21, the purpose is to determine metabolism and pharmacologic actions, side effects associated with increasing doses, and, where possible, early evidence of effectiveness. Participants are typically 20 to 80 people — often healthy volunteers, though oncology and some other Phase 1 trials enroll patients with the target condition because dosing at expected therapeutic levels would otherwise be unsafe in healthy subjects.
Administration burden is disproportionately high relative to trial size. Because the safety profile is essentially unknown, monitoring is intensive — frequently 100% source data verification, on-site or close remote monitoring after every dosing cohort, and dose-escalation decisions that require a safety review committee sign-off between cohorts. Safety reporting is also at its heaviest here: under 21 CFR 312.32, sponsors must report any unexpected fatal or life-threatening suspected adverse reaction to FDA within 7 calendar days, and other serious, unexpected suspected adverse reactions within 15 calendar days — a clock that starts running the moment a single event occurs, regardless of how few participants are enrolled overall.
Phase 2: does it work, and what does it cost in side effects
Phase 2 enrolls a few hundred participants who have the condition the drug is intended to treat. The purpose, per 312.21, is to evaluate effectiveness for a particular indication and to determine common short-term side effects and risks. Many Phase 2 programs are split into 2a (dose-ranging/proof-of-concept) and 2b (dose confirmation) sub-stages, though FDA’s regulation does not formally require this split — it is a common sponsor design choice, not a regulatory category.
This is typically where a trial first needs a full multi-site infrastructure: a coordinating center, a data management plan built to a real statistical analysis plan (see ICH E9 below), and often the first appearance of a Data Safety Monitoring Board with formal charter and interim-analysis authority, rather than the informal safety review common in Phase 1. Monitoring shifts from 100% verification toward risk-based approaches for lower-risk data fields, consistent with the risk-proportionate oversight ICH’s E6(R3) Good Clinical Practice guideline describes, while safety-critical data still gets close review.
Phase 3: pivotal, confirmatory trials
Phase 3 studies exist to generate the “additional information about effectiveness and safety…needed to evaluate the overall benefit-risk relationship” and to provide the basis for physician labeling, per 312.21. These are the pivotal trials a sponsor submits as the primary evidence in an NDA or BLA. Enrollment runs from several hundred to several thousand participants, frequently across many sites and, for large programs, multiple countries.
Trial design at this stage — randomization scheme, control arm, primary and secondary endpoints, the statistical analysis plan — is expected to follow the general framework in ICH E8(R1), General Considerations for Clinical Trials, and the statistical substance in ICH E9, Statistical Principles for Clinical Trials — both of which are pre-specified before enrollment begins because a pivotal trial’s conclusions have to withstand regulatory review. Administratively, Phase 3 is where the full apparatus of modern trial operations shows up: a formal monitoring plan (often centralized/risk-based per ICH E6(R3), supplemented by targeted on-site visits), a functioning DSMB with defined stopping rules, dedicated regulatory affairs support for the eventual NDA/BLA submission, and — for many products intended for FDA-regulated indications — the point at which registration on ClinicalTrials.gov and results reporting obligations under FDAAA 801 most commonly apply. (See CASRAI’s guide to clinical trial registration and reporting compliance for how that obligation is triggered.)
The regulatory transition: from Phase 3 to approval
Completing Phase 3 does not itself change a drug’s regulatory status. The sponsor must submit a New Drug Application (NDA) (for drugs regulated under the Federal Food, Drug, and Cosmetic Act) or a Biologics License Application (BLA) (for biological products regulated under the Public Health Service Act, per 21 CFR 601.2) containing the full nonclinical and clinical data package, including everything gathered across Phase 1–3, before the product can be marketed. After submission, FDA generally has 60 days to decide whether to file the application for review, and a substantive review typically takes 6–10 months once filed. Until that review concludes with an approval, the product remains investigational and every dose administered to a human, in any setting, is still governed by the IND.
Phase 4: post-marketing studies
Phase 4 begins after approval, once the product is on the market. It is no longer governed by the pre-approval IND framework; instead, post-approval studies fall into two categories FDA tracks separately:
- Postmarketing requirements (PMRs) — studies FDA legally requires the sponsor to conduct, under one of four statutory authorities: Section 505(o)(3) of the FD&C Act (added by the Food and Drug Administration Amendments Act of 2007, FDAAA), the Pediatric Research Equity Act (PREA), confirmatory trials required under the Accelerated Approval pathway, or the Animal Rule.
- Postmarketing commitments (PMCs) — studies the sponsor has voluntarily agreed to conduct but that are not independently required by statute or regulation.
Administratively, Phase 4 is qualitatively different from Phase 1–3: enrollment can be large (real-world surveillance cohorts sometimes exceed the entire pre-approval program combined), but the intensity of per-participant monitoring is typically much lower, and safety reporting shifts from the expedited IND-safety-report timelines of 312.32 toward the routine periodic and expedited postmarket adverse-event reporting obligations that apply to an approved product. Staffing shifts accordingly — from a dedicated trial-operations team toward pharmacovigilance and epidemiology functions, often shared across multiple approved products rather than dedicated to one protocol.
How phase classification drives trial administration
The phase a protocol is in is one of the first things a research administration office should check when planning staffing, budget, and monitoring for a new study, because the three biggest operational cost drivers all scale with phase but not in the same direction:
- Staffing intensity per participant is highest in Phase 1 (dedicated coordinators, close physician oversight, dose-escalation safety committees) and declines per-participant through Phase 3 and Phase 4, even as total headcount across a program can grow because absolute enrollment is so much larger.
- Monitoring intensity tends to start near-total (100% source data verification is common in early Phase 1) and shift toward risk-based, centralized monitoring as the safety profile becomes better characterized — the approach ICH E6(R3) formalizes as proportionate to risk rather than uniform across all data.
- Safety reporting burden is heaviest, relative to enrollment size, in Phase 1, where the 7-day/15-day IND safety reporting clocks under 21 CFR 312.32 apply to a small, closely watched population with an unknown safety profile. It doesn’t disappear later — Phase 2 and 3 still carry full IND safety reporting obligations — but the reporting rate per participant is typically lower once common adverse events are already characterized. Phase 4 shifts to a different reporting regime entirely, governed by the approved product’s own postmarket pharmacovigilance requirements rather than IND safety reporting.
Two administrative items commonly change discontinuously at a phase transition rather than scaling gradually: the study almost always needs a new or amended IRB protocol review at each phase transition, since eligibility criteria, dosing, and risk profile typically change; and IMP (investigational medicinal product) supply and manufacturing requirements step up materially from Phase 1 to Phase 3 as batch sizes, packaging/blinding complexity, and GMP documentation scale with enrollment — see CASRAI’s guides to clinical trial supply management and Good Manufacturing Practice for how that scaling works in practice.
“Clinical trial phases,” “clinical trial stages,” and “clinical research phase” — same concept
These three phrasings describe the identical FDA/ICH classification system covered above; none of them refers to a distinct or alternate framework. “Stages” and “phases” are used interchangeably in casual usage, though FDA’s own regulatory text consistently uses “phase.” “Clinical research phase” (singular) typically refers to asking which phase a specific trial is in, rather than the framework as a whole — the answer is one of the five categories in the table above. None of these phrasings should be confused with the distinct NIH-defined clinical trial question (whether a study meets NIH’s four-part definition at all, independent of phase) — see CASRAI’s guide, What Is a Clinical Trial? The NIH Definition Explained, since a Phase 1 study can meet NIH’s definition even though it has no therapeutic goal.
Frequently asked questions
Is Phase 0 the same as Early Phase 1?
Yes. ClinicalTrials.gov’s current phase designation is “Early Phase 1”; “Phase 0” is the older, still commonly used informal name for the same category of exploratory, sub-therapeutic-dose studies conducted under an Exploratory IND.
Does every clinical trial go through all five phases in order?
No. Phase 0 is optional and relatively uncommon; most drug development programs go directly to Phase 1. FDA’s own regulation notes that phases are not necessarily conducted in strict sequence and can overlap — for example, a sponsor may submit an IND covering both Phase 1 and Phase 2 work. Some pathways (accelerated approval, certain oncology indications) also compress or combine phases, such as seamless Phase 1/2 or Phase 2/3 designs.
Do device trials and biologic trials use the same phase system as drug trials?
Biologics use the same Phase 1–4 framework under a BLA rather than an NDA, and the same IND requirement applies before human testing begins. Medical device trials use a different regulatory framework entirely (Investigational Device Exemption, or IDE, under 21 CFR 812, with feasibility and pivotal study categories) rather than the drug/biologic phase system described here.
What triggers the move from one phase to the next?
There is no automatic regulatory trigger — the sponsor decides when accumulated data supports advancing, informed by the safety and efficacy results from the prior phase, and (for later phases) after any required regulatory interactions with FDA, such as an End-of-Phase 2 meeting, which sponsors commonly use to align on Phase 3 design before committing to a pivotal program.
How long does the whole process take, from Phase 1 to approval?
There is no fixed timeline; published industry analyses commonly cite a combined multi-year range across Phase 1 through Phase 3 before an NDA/BLA is even submitted, plus FDA’s own 6–10 month typical review period after filing. Treat any single-figure “average years to approval” claim as a rough industry estimate, not a regulatory benchmark — actual programs vary widely by therapeutic area and trial outcomes.
Does IND safety reporting stop once a drug reaches Phase 3?
No. The 21 CFR 312.32 expedited reporting timelines (7 days for unexpected fatal/life-threatening events, 15 days for other serious unexpected events) apply throughout the entire IND-governed period, from the first Phase 1 dose through the end of Phase 3 — they end only when the IND itself closes, not at an earlier phase transition.
