You have seen the phrase a thousand times. It flashes at the bottom of a TV ad, sits in tiny print on a pharmacy box, and shows up in every news story about a promising new treatment. FDA approved. Two words that most Americans trust without ever asking what stands behind them.
Here is what stands behind them: usually a decade of work, thousands of volunteers, and odds of failure that would scare off most gamblers. The FDA drug approval process is one of the strictest in the world, and understanding it changes how you read a drug ad, a headline, or a clinical trial invitation from your doctor.
This guide walks through every stage, from the first lab experiment to the letter that decides a drug's fate, in plain English.
This article is educational and does not replace medical advice. Talk to a licensed clinician about any treatment decision.
What Is the FDA Drug Approval Process?
The FDA drug approval process is the series of legal and scientific steps a company must complete before it can legally sell a new medicine in the United States. The Center for Drug Evaluation and Research, known as CDER, handles most of that work.
The core question never changes: do the benefits of this drug outweigh its risks for a specific group of patients with a specific condition? Approval is never a blanket statement that a drug is safe. It is a judgment tied to one labeled use, one dose range, and one population.
That distinction matters more than people realize. A drug approved for adults with severe rheumatoid arthritis has not been declared safe for a healthy 22-year-old, or for a child, or for anyone taking it off label.
What the FDA Actually Does and Does Not Do
Most explanations skip the single biggest misunderstanding in this whole topic, so let's fix it up front.
The FDA does not run clinical trials. It does not recruit patients, does not pay for the studies, and does not test drugs in its own laboratories. Pharmaceutical companies, called sponsors, design and fund every study and generate every data point.
What the FDA does is review. Its scientists take the sponsor's raw data, re-analyze it independently, challenge the conclusions, inspect the factories where the drug will be made, and negotiate the wording on the label. Think of the agency as an extremely demanding auditor rather than a research lab.
That model has an obvious tension built into it: the company that profits from approval is the same company producing the evidence. The FDA manages that tension with data audits, site inspections, and the threat of rejection, not by doing the science itself.
The 6 Steps of the US Drug Approval Process
Every prescription medicine sold in America travels the same road:
1. Discovery and preclinical testing in the lab and in animals
2. IND application asking permission to test in humans
3. Clinical trials in three phases with human volunteers
4. NDA or BLA submission containing every scrap of data
5. FDA review and decision, ending in approval or a Complete Response Letter
6. Postmarket surveillance, which never really ends
Start to finish, the trip usually takes 10 years or more. Research from the Tufts Center for the Study of Drug Development put the average cost per approved drug at roughly $2.6 billion, though that figure is debated because it includes more than a billion dollars of estimated capital and time costs rather than cash spent.
Step 1: Discovery and Preclinical Testing
Long before any person swallows a pill, scientists screen thousands of compounds looking for one that hits the right biological target. Most die immediately.
The survivors move to preclinical testing, where researchers study the compound in cell cultures and animals to answer three questions: Is it toxic? At what dose? What does the body do to it, and what does it do to the body?
These studies must follow Good Laboratory Practice regulations, a federal quality standard that governs how the work is documented. Sloppy paperwork here can sink a program years later.
Reality check: industry estimates suggest that out of roughly 250 compounds entering preclinical development, about one eventually reaches pharmacy shelves.
Step 2: The Investigational New Drug (IND) Application
A company cannot simply decide to try its compound on people. It must first file an investigational new drug application with the FDA.
The IND has three main pillars:
1. Animal pharmacology and toxicology data showing the compound is reasonably safe for initial human testing
2. Manufacturing information proving the company can produce consistent, uncontaminated batches
3. Clinical protocols and investigator information describing exactly who will be studied, how, and by whom
The 30-Day Safety Clock and the Role of the IRB
Once the IND lands, a 30-day clock starts. If the FDA raises no objection in that window, the trial may begin. If reviewers spot a serious safety concern, they issue a clinical hold and testing stops until the problem is resolved.
There is a second, independent gatekeeper that patients rarely hear about. Every study site must also get approval from an institutional review board, an ethics committee made up of scientists, clinicians, and community members. The IRB's job is protecting the volunteers, not the science. It reviews consent forms, risk levels, and recruitment methods, and it can shut a study down at its own site regardless of what the FDA says.
Step 3: Clinical Trial Phases 1, 2 and 3
Human testing runs in stages, each one larger and more expensive than the last. Here is how the clinical trial phases compare:
Phase
Participants
Typical Length
Main Question
Roughly How Many Advance
Phase 1
20 to 100
Several months
Is it safe, and at what dose?
About 70%
Phase 2
Up to several hundred
Several months to 2 years
Does it actually work?
About 33%
Phase 3
300 to 3,000
1 to 4 years
Does it work better, and for whom?
About 25% to 30%
Phase 4
Several thousand
Ongoing after approval
What shows up in the real world?
Not applicable
Phase 1 enrolls 20 to 100 people. Phase 3 can reach 3,000.
Phase 1: Is It Safe?
Small groups, often healthy paid volunteers, receive escalating doses under close monitoring. Cancer drugs are the usual exception, since it would be unethical to give a toxic chemotherapy to a healthy person, so those trials enroll patients from the start.
Researchers are watching for side effects, how fast the body absorbs and clears the drug, and where the dose ceiling sits.
Phase 2: Does It Work?
Now the drug goes to people who actually have the condition. This is where the majority of promising compounds die, because a mechanism that looked brilliant in a petri dish frequently does nothing measurable in a human being.
Phase 2 also refines the dose. Too low and the drug fails on efficacy; too high and side effects sink it.
Phase 3: Does It Work Better, and for Whom?
The largest and priciest stage. Hundreds to thousands of patients across many hospitals, usually randomized against either a placebo or the current standard treatment.
Phase 3 exists to catch two things smaller studies cannot: modest real-world benefit, and rare side effects that only appear when you dose thousands of people.
What Percentage of Drugs Survive Clinical Trials?
This is the number that reframes the entire topic, and it is the section worth bookmarking.
Analyses of thousands of drug development programs consistently land in the same range: only about 10% to 14% of drugs that begin Phase 1 testing ever earn FDA approval. Research highlighted by MIT Sloan found close to 14% across a large dataset, while a 2025 analysis in Nature Communications tracking success rates across the 21st century reported figures in a similar band.
The odds swing hard by disease area. Infectious disease vaccines have historically cleared the bar at roughly one in three. Oncology has been far more brutal, with older estimates near 3.4% and more recent work showing improvement into the 8% range as targeted therapies matured.
Stack the phase-level attrition, and the funnel becomes stark:
100 drugs enter Phase 1
About 70 reach Phase 2
About 23 reach Phase 3
Roughly 6 to 7 clear Phase 3
A handful survive FDA review
Of 100 drugs that start Phase 1, only a handful ever reach a pharmacy shelf.
For perspective on the output end, CDER approved 46 novel drugs in 2025, made up of 34 new molecular entities and 12 biologics, according to the agency's annual New Drug Therapy Approvals report. More than half carried orphan designation, meaning they treat rare diseases affecting fewer than 200,000 Americans.
Forty-six approvals. Out of a global pipeline of thousands.
Step 4: The New Drug Application and FDA Review
A single new drug application can run to hundreds of thousands of pages.
A sponsor that survives Phase 3 assembles a new drug application, or NDA, for a conventional chemical drug. Biologic products such as monoclonal antibodies and vaccines go through a biologics license application, or BLA, instead. Generic copies of already approved drugs use a shorter route called an abbreviated new drug application.
These submissions are enormous, often running to hundreds of thousands of pages covering every trial, every adverse event, every manufacturing detail, and proposed label wording.
Who Sits on an FDA Review Team
The application does not go to one bureaucrat. It goes to a multidisciplinary team:
A medical officer, usually a physician, who evaluates clinical efficacy and safety
A statistician who independently re-runs the analyses rather than trusting the sponsor's math
A pharmacologist and toxicologist reviewing the animal data
A chemist examining manufacturing and stability
A clinical pharmacologist assessing dosing and drug interactions
A regulatory project manager coordinating the whole review
Advisory Committees and Facility Inspections
For difficult or high-profile applications, the FDA may convene an advisory committee of outside experts and patient representatives who meet publicly, hear both sides, and vote.
That vote is advice, not law. The clearest recent illustration was aducanumab, the Alzheimer's drug marketed as Aduhelm. The advisory committee voted overwhelmingly against approval in late 2020. The FDA approved it anyway in June 2021 through the accelerated pathway, several committee members resigned in protest, and the manufacturer discontinued the product in 2024.
Separately, FDA investigators physically inspect the plants where the drug will be manufactured. A failed pre-approval inspection can delay an otherwise successful application by a year or more.
The PDUFA Date: 10 Months or 6 Months
Under the Prescription Drug User Fee Act, companies pay substantial fees and the FDA commits to review deadlines in return. The current goals are 10 months for a standard review and 6 months for a priority review, measured from the filing date, which itself falls about 60 days after submission. The agency targets hitting those goals on 90% of applications.
The deadline is called the PDUFA date, and biotech investors watch it the way sports fans watch a playoff schedule. If the sponsor submits a major amendment mid-review, the FDA can extend the clock by three months.
Step 5: The FDA Decision, Approval or Complete Response
At the end of the review, one of two letters goes out.
An approval letter means the drug may be marketed for the indication and population described on the agreed label.
A Complete Response Letter, or CRL, means the application cannot be approved in its current form.
What a Complete Response Letter Really Means
A CRL is not necessarily a death sentence, and this is where casual coverage gets it wrong. The letter lists the specific deficiencies and what the company must do to fix them.
Sometimes the fix is scientific, such as an additional trial in a different population. Often it is far more mundane. Manufacturing problems at a third-party plant, an unresolved inspection finding, or a labeling dispute have all triggered CRLs for drugs whose clinical data was never in question. Plenty of drugs receive a CRL, resubmit, and get approved on the second pass.
Why FDA Started Publishing Rejection Letters
For decades, CRLs were confidential. Companies would announce a "regulatory setback" in a vague press release, and outsiders had no way to know whether the FDA had rejected the drug over a failed trial or a paperwork problem.
That changed sharply:
1. July 10, 2025: the FDA published more than 200 previously confidential CRLs on its openFDA platform, covering applications submitted between 2020 and 2024 that had ultimately been approved.
2. September 4, 2025: the agency released 89 more, this time tied to pending or withdrawn applications, and said it would begin posting letters closer to real time.
3. April 2026: new releases were paused while the agency reviewed a citizen petition challenging the legal basis for disclosing what companies argue is confidential commercial information.
Whichever way that fight resolves, the archive already published on openFDA is a genuinely new window into how the agency thinks, and it is the single best free resource for anyone who wants to see FDA reasoning in its own words.
Step 6: Phase 4 and Postmarket Surveillance
Approval is a checkpoint. Monitoring continues for as long as the drug is on the market.
Approval is a checkpoint, not a finish line. A Phase 3 trial of 3,000 people simply cannot detect a side effect that strikes one patient in 20,000.
Once a drug reaches the market and millions of people take it, several systems keep watching:
1. Phase 4 studies, sometimes voluntary and sometimes required as a condition of approval
2. FAERS, the FDA Adverse Event Reporting System, which collects reports from clinicians, patients, and manufacturers
3. Label updates, including new warnings or interaction data as evidence accumulates
4. Boxed warnings, the most serious safety alert the FDA can require on a label
5. REMS, a Risk Evaluation and Mitigation Strategy that can restrict who may prescribe or dispense a drug and require patient monitoring
When the postmarket picture turns bad enough, drugs come off the market. It is uncommon, but it happens, and that possibility is a feature of the system rather than a failure of it.
Faster Routes: The 4 FDA Expedited Pathways
Not every drug waits its turn. Congress built four programs for serious conditions, and they are routinely confused with one another.
Pathway
What It Actually Changes
When It Applies
Fast Track
More frequent FDA meetings, eligibility for rolling review
Serious condition plus potential to fill an unmet need
Breakthrough Therapy
Intensive FDA guidance, senior staff involvement
Early clinical evidence of substantial improvement over available therapy
Priority Review
Shortens the review clock from 10 months to 6
Significant improvement in safety or effectiveness
Accelerated Approval
Lowers the evidence bar to a surrogate endpoint
Serious condition with an unmet need and a marker reasonably likely to predict benefit
Three pathways change speed. Only one changes the evidence standard.
Three of these are designations that change how a drug is developed or how fast it is reviewed. Only accelerated approval changes what evidence is required for approval itself. That is the crucial distinction, and it gets blurred constantly in news coverage.
The pathways also stack. A single drug can hold breakthrough designation, get priority review, and be approved under accelerated approval all at once.
The Catch With Accelerated Approval
Accelerated approval lets the FDA clear a drug based on a surrogate endpoint, a lab measurement thought to predict real benefit, such as tumor shrinkage standing in for longer survival.
The trade is explicit: the company must run confirmatory trials afterward to prove the actual clinical benefit exists. If those trials fail, the drug is supposed to come off the market.
That is not theoretical. Makena, a preterm birth drug approved on an accelerated basis in 2011, was withdrawn in April 2023 after its confirmatory trial failed to show benefit. The gap between approval and withdrawal was 12 years, during which the product was widely prescribed to pregnant women.
Congress tightened this in the 2022 Food and Drug Omnibus Reform Act, giving the FDA authority to require confirmatory trials be underway before approval and to pull drugs faster when they fail. So when you read that a drug received accelerated approval, read it as conditional and provisional, not settled.
Approved vs Cleared vs Authorized: What the Labels Mean
Four different legal statuses get used interchangeably in conversation, and they mean very different things:
1. Approved means the FDA reviewed evidence of safety and effectiveness and said yes. This covers prescription drugs, biologics, and the highest-risk medical devices.
2. Cleared applies to medical devices that went through the 510(k) route by demonstrating they are substantially equivalent to a device already on the market. No new clinical trial is necessarily involved. Most everyday devices are cleared, not approved.
3. Authorized refers to an Emergency Use Authorization, a temporary status used during a declared public health emergency when there is no adequate approved alternative. It is a lower bar, and it expires.
4. Granted describes the De Novo pathway, used for novel low to moderate-risk devices with no existing equivalent.
If a supplement bottle or wellness gadget implies FDA approval, check the wording carefully. Dietary supplements are not approved by the FDA at all. Manufacturers are responsible for their own safety substantiation, which is a fundamentally different arrangement from what prescription medicines go through.
What This Means If You Are Offered a Clinical Trial
Understanding the pipeline is useful in the abstract. It becomes personal the moment a doctor mentions a study.
More than 400,000 studies are registered on ClinicalTrials.gov, the federal registry run by the National Library of Medicine. You can search it yourself by condition, location, and recruitment status, and every entry lists eligibility criteria and a contact.
Before agreeing to anything, know your protections and ask the right questions:
Informed consent is a process, not a signature. The team must explain purpose, procedures, risks, and alternatives in language you understand, and you can take the document home to read.
You may withdraw at any time, for any reason, without losing your regular medical care.
An IRB has reviewed the study, and you can ask which one.
Ask which phase it is. A Phase 1 trial is primarily testing safety, not treating you.
Ask about placebo. Will you know? What happens if you get the placebo arm?
Ask about cost. Who pays for the study drug, the extra scans, the travel?
Phase 1 tests safety and dosing in 20 to 100 people. Phase 2 tests whether the drug works in up to several hundred patients. Phase 3 confirms benefit in 300 to 3,000 patients. Phase 4 monitors safety after approval in real-world use.
Development typically runs 10 years or more from discovery to market. FDA review itself is faster, with goals of 10 months for standard review and 6 months for priority review, counted from the filing date about two months after submission.
Roughly 10% to 14% of drugs entering Phase 1 eventually win approval. Success varies widely by disease, with infectious disease vaccines historically performing far better than oncology drugs.
The drug was approved based on a surrogate endpoint likely to predict benefit, not on proven clinical outcomes. The company must run confirmatory trials, and the drug can be withdrawn if those trials fail, as happened with Makena in 2023.
No. Companies design, fund, and run all trials. The FDA reviews and independently re-analyzes the submitted data, inspects manufacturing facilities, and controls the final label.
Monitoring continues through Phase 4 studies, the FAERS adverse event database, label changes, boxed warnings, and REMS restrictions. Drugs with serious postmarket problems can be pulled from the market.
The Bottom Line
The FDA drug approval process is slow, expensive, and unglamorous, and those are largely features rather than bugs. Ninety percent of drugs that reach human testing never make it, which means the system's main product is rejection.
Knowing the stages gives you a sharper filter. You can tell an accelerated approval from a full one, read a Complete Response Letter headline without panicking, spot the difference between a cleared device and an approved drug, and ask better questions if a trial is ever offered to you or someone you love.
Next time you see "FDA approved," you will know exactly how much work those two words are carrying.
Found this useful? Share it with someone who is weighing a clinical trial or trying to make sense of a drug headline, and tell us in the comments which part of the process surprised you most.
Published by Certify Me USA
Certify Me USA Editorial Team
Certify Me USA is a US-focused resource covering healthcare, workplace safety, and professional certification. This guide is based on analysis of FDA and CDER published data, PDUFA performance goals, and peer-reviewed research on clinical development success rates.
Make Sense of the FDA Drug Approval Process
Get clear, source-backed guides on US healthcare rules, clinical research, and certification delivered as we publish them.