If you are trying to make sense of the drug development process stages, here is the short version: a target is chosen, molecules are screened and refined, safety is tested, humans are studied, and a submission follows.
Each stage is explained below in plain language, with links to the regulator and research pages the details come from.
| Item | Value |
|---|---|
| Historical time from discovery to approval | Approximately 10 to 15 years |
| Typical cost reported in a 2025 review | Often exceeding USD 1 to USD 2 billion |
| Candidates entering Phase I that are approved | Estimates suggest fewer than 1 in 10 |
| Overall success rate, non-cancer rare disease drugs | 25% |
| Overall success rate, projects that included biomarkers | 26% |
The figures come from a 2025 systematic review in Pharmaceuticals and a 2016 Nature Reviews Drug Discovery news item. They can change as new data appear, so check the sources before you rely on them.
Drug development process stages at a glance
Drug development is best pictured as a funnel. Many ideas enter at the top, and only a few molecules reach the end. The stages overlap in practice, and teams often loop back when a result is disappointing.
For the broader background, our guide on what is drug discovery explains how the field is organised. This article walks through the stages in order and flags where things tend to go wrong.
- Target identification and validation
- Assay development, screening and hit identification
- Lead optimization and candidate selection
- Preclinical testing and preparation for human studies
- Clinical trials in people
- Submission to regulators and review
Stage one: choosing and validating a target
A target is usually a protein or pathway thought to be involved in a disease. One review of early drug discovery notes that the idea for a target can come from a variety of sources including academic and clinical research and from the commercial sector.
Validation means gathering evidence that changing the target might plausibly affect the disease. The same review adds that it may take many years to build up a body of supporting evidence before selecting a target for a costly drug discovery programme. A weak target can lead to failure many years later, which is why this stage matters.
Stages two and three: from hits to a lead candidate
Once a target is chosen, researchers build a laboratory test, called an assay, and use it to screen large libraries of molecules. Molecules that show activity are called hits. The review above describes the preclinical path as running from target identification and validation, through assay development, high throughput screening, hit identification, lead optimization and finally the selection of a candidate molecule for clinical development.
Lead optimization is where chemists adjust a promising molecule to improve its properties while keeping what works. Our explainer on hit to lead optimization covers this step in more detail.
A computational prediction does not replace laboratory and human evidence.
Preclinical work and the step into people
Developers also prepare the paperwork that regulators need before human studies can begin. The 2025 review mentions a transitional phase between preclinical work and early human studies, often associated with IND (Investigational New Drug) enabling activities.
Exact requirements differ between regulators and between types of medicine. Check the current guidance from the relevant agency rather than relying on a general summary.
This stage is also where manufacturing questions appear. Problems found here can delay a programme or end it before any person receives the candidate.
Researchers sometimes use biomarkers, measurable signs in the body, to follow how a candidate behaves. The Nature Reviews Drug Discovery item above reported that projects including them had a higher overall success rate, though that is an association and not proof of cause.
Clinical trials and submission
Human studies are run in phases that ask different questions, moving from small safety-focused studies to larger ones that look at whether a candidate appears to work. We cover the details in clinical trial phases explained.
When a developer believes it has enough evidence, it compiles the results into a marketing submission. Regulators then review the data and decide whether the evidence supports authorisation. A review can end with a request for more data, which sends the developer back to the clinic or the laboratory.
How long it takes and how often it fails
The 2025 review says that historically, it takes approximately 10 to 15 years to develop a new therapeutic agent from initial discovery to regulatory approval, with costs often exceeding USD 1 to USD 2 billion. It also states that estimates suggest fewer than 1 in 10 drug candidates entering Phase I clinical trials are ultimately approved.
Success rates vary. A 2016 Nature Reviews Drug Discovery news item reported considerable variation by therapeutic area and drug modality. It put the overall success rate of non-cancer rare disease drugs at 25% and of projects that included biomarkers at 26%.
Researchers also debate why so many candidates fail. A 2022 paper titled Why 90% of clinical drug development fails and how to improve it? argues that current drug optimization emphasises potency and specificity while overlooking how much of a drug reaches diseased versus normal tissue. That is one proposal, not settled fact.
- Numbers depend on the dataset, the years covered and how success is defined.
- Older figures may not match current practice.
- Averages say little about any single candidate.
What computational tools can and cannot do
A prediction is a hypothesis that still has to be tested at the bench.
The hardest questions are usually biological.
- Early results may not hold up in animals or people.
- Human biology is variable, so averages can hide important differences.
Reading development claims with care
News about a candidate usually describes one stage only. A promising laboratory result sits far from a finished medicine, and most candidates at that point never reach patients. When you read such a story, ask which stage the work is at and what kind of evidence it rests on.
It also helps to check who ran the study, whether it was peer reviewed and whether the numbers come from people or from cells and animals. Treat company statements as claims to verify, not as findings.
This site is general education only. Decisions about any medicine or condition belong with a doctor or other qualified health professional.
Frequently asked questions
What are the main drug development process stages?
Show answer
They run from target identification and validation, through screening, lead optimization and preclinical testing, to clinical trials and a regulatory submission. The stages overlap and teams often repeat earlier steps.
How long does drug development take?
Show answer
A 2025 systematic review states that it historically takes approximately 10 to 15 years from initial discovery to regulatory approval. Individual programmes can be shorter or longer.
Do most drug candidates fail?
Show answer
That is what the published estimates suggest. The same review says fewer than 1 in 10 candidates entering Phase I are ultimately approved.
Can AI shorten the process?
Show answer
AI tools are being applied at several stages, but evidence of their effect is still being gathered. They do not remove the need for laboratory and clinical testing.
This article is for information only. It is general education about science and the drug industry, and it is not medical, treatment or investment advice.
