Last reviewed · How we verify
Phase 1 vs Phase 2 vs Phase 3 clinical trials
Phase 1 — first-in-human safety
Phase 1 trials are the first time a drug is administered to humans. Goal: establish maximum tolerated dose, characterise pharmacokinetics (how the body processes the drug), and identify acute safety signals. Typical enrolment: 20–80 participants (healthy volunteers for non-oncology; patients for oncology where dosing in healthy people is unsafe). Duration: 6 months to 2 years. Success rate to Phase 2: ~65%.
Phase 2 — proof of concept
Phase 2 establishes whether the drug actually works in patients with the target disease. Endpoint is typically a surrogate marker — tumour response in oncology, HbA1c in diabetes, ACR20 in rheumatology. Phase 2 trials are randomised vs placebo or standard of care, enrolling 100–500 patients. Duration: 1–3 years. Phase 2 is the biggest attrition step: only ~30% of Phase 2 candidates advance.
Phase 3 — pivotal evidence
Phase 3 is the confirmatory study used for FDA approval. Endpoint must be clinically meaningful — overall survival, hospitalisation reduction, disease progression. Trials are large randomised controlled studies (typically 1,000–5,000 patients), often two separate trials to provide replication, vs placebo or best standard of care. Duration: 3–5 years. Success rate: ~58%.
After Phase 3 — Phase 4
After FDA approval, sponsors continue post-marketing surveillance via Phase 4 trials (sometimes called post-authorisation safety studies, PASS). These can be voluntary or FDA-mandated as a condition of approval. They monitor real-world effectiveness, long-term safety, rare adverse events, and outcomes in subpopulations excluded from Phase 3.
Specialised trial designs
Modern oncology often uses adaptive designs that allow Phase 2 + Phase 3 to be combined ("Phase 2/3"), basket trials (single drug across multiple cancers with the same target), umbrella trials (multiple drugs in one cancer), and platform trials (open-ended frameworks adding new drugs over time). These compress timelines but make trial comparisons harder.
Where to track active trials
ClinicalTrials.gov (US-mandated registry) is the primary source. Drug Landscape ingests this plus EU Clinical Trials Register, UK ISRCTN, China CTR, India CTRI. Each trial has a NCT identifier (NCT04812548). Our /trial/[nct] page shows the design, endpoints, enrollment, and any posted results.
FAQ
How long does each phase take?
Phase 1: 6 months to 2 years (typically 1 year). Phase 2: 1–3 years. Phase 3: 3–5 years. Total: 6–10 years from first dose to approval filing, on top of ~3–6 years of preclinical work.
Why do most drugs fail in Phase 2?
Phase 1 measures safety + PK — relatively easy to achieve. Phase 2 measures efficacy — this is where most drugs reveal they don't actually work, or work only in subgroups. Phase 2 attrition has been stable around 70% for decades despite enormous investment in target validation.
Are Phase 3 readouts always pivotal?
Mostly yes for the primary indication. But indication-expansion trials (e.g. Keytruda in a 15th cancer type) are also Phase 3 and follow similar regulatory pathways. So a single drug can have many Phase 3 readouts over its lifecycle.
Why do clinical trials cost so much?
Phase 3 trials average $20–50M but can reach $200M+ for large indications. Costs come from: patient recruitment + clinical site fees, data management, regulatory + monitoring overhead, comparator drug supply, biostatistics, regulatory submission. Cell + gene therapy trials are especially expensive due to manufacturing complexity.