6 min read
What percentage of clinical trials fail?
Dr. Matic Broz Computational chemist
Table of contents
About nine in ten drug candidates that enter human clinical testing do not become approved medicines. The exact percentage depends on what is being counted: a molecule, a drug-indication program, a company-sponsored path, or a regulatory filing.
This distinction matters. A "failed clinical trial" can mean a trial missed its endpoint, a sponsor stopped a program, a drug never advanced to the next phase, or a candidate reached the FDA and was not approved.
What percentage of drugs fail clinical trials?
Published estimates cluster around the same answer: most drugs entering clinical development fail before approval. In the BIO 2011-2020 benchmark, the Phase 1-to-approval likelihood was 7.9%, which implies a 92.1% failure rate for company-sponsored FDA registration-enabling programs.[1]
A newer Citeline/Biomedtracker analysis of 2014-2023 phase transitions put the Phase 1-to-approval likelihood at 6.7%, implying a 93.3% failure rate. A large academic study using 406,038 clinical-trial data entries from 2000-2015 estimated a higher 13.8% Phase 1-to-approval success rate, or an 86.2% failure rate.[2][3]
| Estimate | What it counts | Success rate | Failure rate |
|---|---|---|---|
| Citeline/Biomedtracker, 2014-2023 | New Phase 1 drug LOA | 6.7% | 93.3% |
| BIO/QLS/Informa, 2011-2020 | Company-sponsored FDA programs | 7.9% | 92.1% |
| Wong, Siah, and Lo, 2000-2015 | Programs reconstructed from trial data | 13.8% | 86.2% |
Sources: Citeline/Biomedtracker, BIO/QLS/Informa, and Wong et al. Failure rates are calculated as 100% minus the reported success rate.[2][1][3]
A 2025 Nature Communications analysis adds why the spread is not just noise: success rates change when the unit is a molecule instead of a drug-indication clinical development program. The authors reported large variation across definitions, diseases, modalities, and time windows, with recent molecule-level estimates higher than drug-indication program estimates.[4]
What percentage fail by clinical trial phase?
Phase 2 is the biggest hurdle: in the BIO 2011-2020 benchmark, 71.1% of programs failed to advance from Phase 2 to Phase 3. Phase 3 is not safe either; 42.2% failed to move from Phase 3 to an NDA/BLA filing, and 47.6% failed to reach approval when the whole Phase 3-to-approval path is counted.[1]
NDA/BLA means New Drug Application or Biologics License Application, the FDA filing step after successful late-stage development.
| Starting point | Outcome | Success rate | Attrition |
|---|---|---|---|
| Phase 1 | Advances to Phase 2 | 52.0% | 48.0% |
| Phase 2 | Advances to Phase 3 | 28.9% | 71.1% |
| Phase 3 | Files an NDA or BLA | 57.8% | 42.2% |
| Phase 3 | Ultimately receives FDA approval | 52.4% | 47.6% |
| NDA/BLA filing | Receives FDA approval | 90.6% | 9.4% |
The table uses BIO's 2011-2020 phase-transition data; attrition rates are calculated as 100% minus the reported success or likelihood-of-approval rates.[1]
The phase labels matter. FDA describes Phase 1 as early human safety and dosing work, Phase 2 as preliminary efficacy and short-term safety in patients, and Phase 3 as larger studies meant to evaluate benefit-risk and support labeling. That is why Phase 2 often filters aggressively before companies commit to much larger Phase 3 studies.[5]
The newer 2014-2023 Citeline analysis tells the same story with slightly lower rates: Phase 1 success of 47%, Phase 2 success of 28%, Phase 3 success of 55%, and post-filing approval of 92%.[2]
Some search summaries describe Phase 2 as a 30% to 40% failure point. In these phase-transition datasets, that range is closer to the Phase 2 success rate: BIO reports 28.9% advancing to Phase 3, and Citeline reports 28%. If failure means not advancing to Phase 3, Phase 2 attrition is about 71% to 72%.[1][2]
Why do drugs fail in clinical trials?
Most clinical drug failures are caused by lack of efficacy or safety problems, not by paperwork alone. In a JAMA Internal Medicine study of 640 novel therapeutics in late-stage development, 344 failed; among the failed agents, 57% failed for inadequate efficacy, 17% for safety, and 22% for commercial or other reasons.[6]
A separate review of Phase 2 and Phase 3 failures from 2013-2015 found a similar pattern: 52% of failures were due to lack of efficacy, 24% to safety, 15% to strategic reasons, 6% to commercial reasons, and 3% to operational reasons.[7]
| Reason | Meaning | Share |
|---|---|---|
| Lack of efficacy | Not enough clinical benefit | 52% to 57% |
| Safety or toxicity | Harm outweighs benefit | 17% to 24% |
| Commercial, strategic, or operational | Sponsor stops for market, portfolio, recruitment, or trial reasons | About 22% to 24% |
Sources: Hwang et al. and Harrison. Categories differ slightly between studies, so the table groups comparable reasons rather than averaging them.[6][7]
The broader "90% failure" literature also points to problems that start before the trial. Sun and colleagues summarize clinical-development failures as mainly lack of efficacy (40% to 50%), unmanageable toxicity (about 30%), poor drug-like properties (10% to 15%), and commercial or strategic issues (about 10%).[8]
Can clinical trial success rates be improved?
Yes, but the strongest evidence points to better target choice and better patient selection, not a single fix to trial operations. In the BIO 2011-2020 benchmark, programs using patient preselection biomarkers had a 15.9% Phase 1-to-approval likelihood, compared with 7.6% for programs without those biomarkers.[1]
The difference was largest in Phase 2, where biomarker-selected programs advanced 46.3% of the time versus 28.3% for programs without preselection biomarkers. Wong, Siah, and Lo found a similar signal in the 2005-2015 subset: biomarker patient-selection trials had nearly twice the overall probability of success, 10.3% versus 5.5%.[1][3]
Target biology matters too. A 2024 Nature analysis found that drug mechanisms with human genetic support were 2.6 times more likely to succeed than mechanisms without it, with the effect strongest when the causal gene evidence was clearer.[9]
These are improvements in odds, not guarantees. They help explain why the best answer is not just "90% fail", but "most fail, especially in Phase 2, and the failures are concentrated where human biology was not predicted well enough." The same attrition shapes drug discovery trends and is also why drug development cost estimates count failed programs, not just the one medicine that reaches patients.
FAQ
Is it true that 90% of clinical drug development fails?
Yes, it is largely true. Approximately 90% of drug candidates that enter clinical trials fail to reach the market. This massive attrition rate, which even climbs to over 92% in certain therapeutic areas like oncology and neurology, is why developing a single successful medicine often takes 10 to 15 years and costs billions of dollars.
The supporting benchmarks are the BIO 2011-2020 success rates for all programs, oncology, and neurology, plus ASPE's 10- to 15-year development-timeline estimate.[1][10]
ProteinIQ. "What percentage of clinical trials fail?" Updated July 1, 2026. Accessed [your access date]. https://proteiniq.io/guides/clinical-trial-failure-rate
Sources▼
- Clinical Development Success Rates and Contributing Factors 2011-2020 BIO, QLS Advisors, and Informa Pharma Intelligence · 2021. https://go.bio.org/rs/490-EHZ-999/images/ClinicalDevelopmentSuccessRates2011_2020.pdf
- Why Are Clinical Development Success Rates Falling? Norstella · July 1, 2026. https://www.norstella.com/insight/why-are-clinical-development-success-rates-falling/
- Estimation of clinical trial success rates and related parameters Biostatistics · 2019. https://academic.oup.com/biostatistics/article/20/2/273/4817524
- Dynamic clinical trial success rates for drugs in the 21st century Nature Communications · 2025. https://www.nature.com/articles/s41467-025-64552-2
- Step 3: Clinical Research U.S. Food and Drug Administration · July 1, 2026. https://www.fda.gov/patients/drug-development-process/step-3-clinical-research
- Failure of Investigational Drugs in Late-Stage Clinical Development and Publication of Trial Results JAMA Internal Medicine · 2016. https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2565686
- Phase II and phase III failures: 2013-2015 Nature Reviews Drug Discovery · 2016. https://www.nature.com/articles/nrd.2016.184
- Why 90% of clinical drug development fails and how to improve it? Acta Pharmaceutica Sinica B · 2022. https://www.sciencedirect.com/science/article/pii/S2211383522000521
- Refining the impact of genetic evidence on clinical success Nature · 2024. https://www.nature.com/articles/s41586-024-07316-0
- Examination of Clinical Trial Costs and Barriers for Drug Development HHS Office of the Assistant Secretary for Planning and Evaluation · 2014. https://aspe.hhs.gov/reports/examination-clinical-trial-costs-barriers-drug-development
