Key points
- Only 28.9% of Phase 2 programs advanced to Phase 3 in the 2011–2020 BIO industry analysis, the lowest rate of any phase.
- Combining that with the later-stage rates in the same report gives about a 15% chance of approval for a typical drug entering Phase 2.
- That figure ranges from roughly 10% in oncology and cardiovascular disease to about 34% in hematology.
- Programs that selected patients with biomarkers had roughly double the odds from Phase 2: about 30%, compared with about 15% without biomarkers.
- A registry-based study using a different method estimated a higher Phase 2-to-approval rate of 35.1%, a reminder that these figures depend on how they are calculated.
Phase 2 is where most experimental drugs fail. It is usually the first time a drug is properly tested for effect in people with the disease, and it is where the scientific assumptions made during discovery meet reality. For a typical drug entering Phase 2, the historical odds of eventual FDA approval are roughly 1 in 7.
The headline number
The 2021 BIO, Informa Pharma Intelligence, and QLS Advisors analysis of 2011–2020 programs reported these transition rates for the stages after Phase 2 begins:
| Step | Success rate |
|---|---|
| Phase 2 → Phase 3 | 28.9% |
| Phase 3 → application submitted | 57.8% |
| Application → approval | 90.6% |
Multiplying the three rates together gives the likelihood that a drug entering Phase 2 is eventually approved:
28.9% × 57.8% × 90.6% ≈ 15.1%
The report publishes this figure directly in its likelihood-of-approval table (Figure 5b): 15.1% from Phase 2, alongside 7.9% from Phase 1.
Almost all of the attrition happens in the first step. More than 7 in 10 programs that enter Phase 2 never reach Phase 3.
By disease area
The report's likelihood-of-approval table shows how much the odds vary by disease area:
| Disease area | Phase 2 → Phase 3 | Phase 2 → approval |
|---|---|---|
| Hematology | 48.1% | 34.4% |
| Metabolic | 45.0% | 25.0% |
| Infectious disease | 38.4% | 22.8% |
| Autoimmune | 31.4% | 19.3% |
| All indications | 28.9% | 15.1% |
| Psychiatry | 26.8% | 13.8% |
| Respiratory | 21.9% | 13.5% |
| Neurology | 26.8% | 12.3% |
| Oncology | 24.6% | 10.8% |
| Cardiovascular | 21.0% | 9.6% |
Source: BIO, Informa Pharma Intelligence, QLS Advisors (2021), Figures 2 and 5b. The report does not publish which conditions it assigns to each disease area.
Cardiovascular disease and oncology sit near the bottom for different reasons. Cardiovascular trials often need to show an effect on hard outcomes, such as heart attacks or deaths, in very large populations. Oncology has a huge number of programs, and many of them test new mechanisms in biologically varied tumors.
The biomarker effect
One of the strongest patterns in the data concerns patient preselection biomarkers: tests used to enroll only the patients most likely to respond, such as those with a specific genetic mutation.
| Programs | Phase 2 → Phase 3 | Phase 2 → approval (calculated) |
|---|---|---|
| With preselection biomarkers | 46.3% | 30.3% |
| Without biomarkers | 28.3% | 14.6% |
Source: BIO, Informa Pharma Intelligence, QLS Advisors (2021). Biomarker programs made up about 6% of transitions in the analysis.
Selecting the right patients roughly doubles the odds. It reflects a deeper point: Phase 2 often fails not because a drug does nothing, but because its effect is diluted across patients who were never going to respond.
A different method, a different answer
Not every analysis agrees on the level. Wong, Siah, and Lo (Biostatistics, 2019) used trial-registry data from 2000 to 2015 and a "path-by-path" method that follows each drug–indication program through its full history. They estimated a Phase 2 → Phase 3 rate of 58.3% and a Phase 2 → approval rate of 35.1%, more than double the BIO figure.
Both studies are careful, and both are widely cited. The gap shows that success rates are not natural constants. They depend on how a "program" is defined, which data source is used, and how silent or unreported programs are treated. See How Are Drug Approval Probabilities Calculated?
Why Phase 2 is so hard
- It is the first real efficacy test. Phase 1 shows a drug is tolerable. Phase 2 asks whether it helps.
- Animal models often mislead. Effects seen in animals frequently do not translate to human disease.
- Dose is still uncertain. A dose chosen for safety in Phase 1 may be too low to work.
- Trials are small. Phase 2 studies usually enroll up to a few hundred patients, enough to detect large effects but not modest ones.
What this means for a specific drug
A historical average describes a group of programs, not an individual drug. Two Phase 2 programs can have very different prospects. One might have a randomized result, a validated target, and a biomarker-selected population, while another has only an uncontrolled signal in a broad population. What Are the Chances of FDA Approval After Phase 2? explains the factors that move an individual program above or below the average.
Sources
- Biotechnology Innovation Organization, Informa Pharma Intelligence, QLS Advisors. Clinical Development Success Rates and Contributing Factors 2011–2020 (2021).Industry report
- Wong CH, Siah KW, Lo AW. Estimation of clinical trial success rates and related parameters. Biostatistics. 2019;20(2):273–286.Peer-reviewed
- Hay M, Thomas DW, Craighead JL, Economides C, Rosenthal J. Clinical development success rates for investigational drugs. Nature Biotechnology. 2014;32:40–51.Peer-reviewed
- U.S. Food and Drug Administration. Step 3: Clinical Research.Regulator
This article is for education. It is not medical advice and not investment advice. Figures are drawn from the sources above and describe historical averages, not the prospects of any particular drug. See our editorial standards and disclaimer.