How Low Can They Go? Clinical Trial Success Rates

New data published in Nature Communications (October 2025) — derived from 20,398 clinical trial pipelines covering 9,682 unique molecular entities over two decades — confirm something terrible that the industry doesn’t like to admit: Clinical trial success rates are not just terribly low, they’re declining.

Image

What is the Rate of Clinical Trial Success?

What is the real likelihood that a drug entering clinical trials will ever really reach patients? And why can no one in the pharma industry offer a straightforward answer?

Reported clinical trial success rates have varied from as low as 7% to as high as 20%, depending on the methodology and dataset used.

This creates persistent confusion across the industry.

The wide range reflects genuine differences in how researchers define a “development path,” the time windows applied, and whether data comes from public registries, commercial databases, or company disclosures.

A landmark 2019 MIT biostatistics review offers one of the most rigorous estimates, drawing on 406,038 clinical trial entries for over 21,143 compounds between 2000 and 2015.

The overall probability of drug development success — from Phase I to approval — was estimated at approximately 13.8% for industry-sponsored trials. The study found significant variation by disease area: Oncology, for example, carried an overall success rate of just 3.4% in their sample, against a previously cited figure of 5.1%.

A 2025 Nature Communications analysis extends this picture further, revealing that overall clinical success rates have been declining since the early 2000s, though they appear to have recently plateaued and may be beginning a modest recovery.

Critically, the latest study introduces a dynamic measurement framework (the ClinSR platform at ClinSR.org) that enables real-time tracking of success rates rather than relying on static historical windows.

Image

View interview Is AI Quietly Creating a GDPR Crisis in Clinical Research?

Is AI Quietly Creating a GDPR Crisis in Clinical Research?

Generative AI is transforming clinical trials, but a failure to distinguish between pseudonymised and anonymised data is creating costly regulatory gaps.

Image

What Do Phase-by-Phase Transition Rates Look Like?

Understanding clinical trial and drug development programme success rates requires examining where, precisely, the pipeline attrition occurs.

The MIT Biostatistics study found that the probability of advancing from Phase I to Phase II was approximately 63.2%, from Phase II to Phase III approximately 30.7%, and from Phase III to approval approximately 58.1%.

These figures give a compound probability, multiplied across all transitions, that brings the overall likelihood of approval to the low teens.

A 2025 analysis cited by IntuitionLabs updates these phase-specific figures: Phase II success rates have declined to approximately 28%, while Phase I and Phase III hover near 47% and 55%, respectively.

The Phase II attrition rate is particularly consequential because it is at this stage that drugs begin to reveal whether they actually work in patients. This is also the most costly point in development to encounter failure.

Success Rate by Disease Area and Drug Modality 

There is enormous variation in success rate by both disease area and drug modality.

The latest 2025 Nature analysis shows substantial divergence across disease categories, developmental strategies, and drug modalities.

Oncology consistently posts the lowest success rates across all analyses.

The MIT study found a 3.4% end-to-end success rate for oncology compounds — far below the aggregate.

The oncology figure improved considerably in 2014 and 2015 (rising to 8.3% in 2015), driven largely by the emergence of immuno-oncology and biomarker-driven patient selection strategies. However, it also declined again in subsequent years as the post-checkpoint inhibitor excitement gave way to harder target biology realities.

Drug modality is an increasingly important factor for clinical success.

The Nature Communications data show the share of small molecular drugs in clinical pipelines falling from 66.3% at the start of the 2000s to 46.6% by 2024, while antibody-related drugs have grown from 10.8% to 19.5% of all clinical-stage entities.

Biologics, gene therapies, and RNA-based modalities carry different risk profiles — and, in select categories, meaningfully higher success rates — a dynamic that reshapes how portfolio risk should be treated.

Image

A Notable Anomaly: Drug Repurposing Underperforms

One counterintuitive finding from the 2025 study deserves particular attention: Repurposed drugs have had unexpectedly lower success rates than newly developed compounds in recent years.

The conventional assumption has been that drugs already proven safe in human patients should carry lower clinical risk when redirected to new indications.

This is directly challenged by the data, suggesting that the biological rationale for repurposing candidates may be weaker than often claimed, or that repurposing programmes are frequently conducted without the disciplined patient selection and endpoint design that new compounds receive.

Anti-COVID-19 drugs represent a separate anomaly: The analysis identifies an extremely low clinical success rate for this category, likely reflecting the unprecedented speed and volume of trials conducted under emergency conditions rather than structured efficacy-driven development.

What Clinical Trials Success Rate Numbers Mean for Pharma R&D

For VP- and C-suite-level pharma R&D leaders, clinical trial success rates are the denominator in every pipeline investment decision.

A 13% to 14% overall success rate means that, on average, a portfolio must carry seven or more clinical-stage candidates to expect a single approval. That calculation underpins the entire logic of therapeutic area focus, licensing strategy, and capital allocation.

Several evidence-based levers have been shown to meaningfully improve the probability of success.

Trials that incorporate biomarkers for patient selection carry higher overall success rates than those without, a finding consistent across multiple major analyses.

First-in-class programmes in well-validated mechanisms outperform me-too strategies over long timelines.

And, critically, the therapeutic area and modality selection made at the portfolio level shapes the statistical baseline before a single trial is designed.

The introduction of dynamic, continuously updated success rate tracking (as offered by ClinSR.org) now enables R&D leadership to benchmark their own pipeline assumptions against real-time industry data rather than five-year-old historical averages.

Pharmatica provides the strategic intelligence that connects clinical trial data to real-world R&D decisions, enabling pharmaceutical leaders to benchmark performance, evaluate risk, and allocate capital with the clarity that static industry reports cannot provide.

Pharmatica: Insight. Connection. Impact.

Image