Last updated: October 31, 2026 · Dr Nidhi Joshi
In short: The drug discovery process is the long road from a disease question to an approved medicine. The main stages of drug discovery and development are: find and validate a target, screen for hits, optimise a lead, test it in preclinical studies, run clinical trials in people, then pass regulatory review and post-market safety monitoring.
What is the drug discovery process?
Every medicine in your cupboard began as a question: what goes wrong in this disease, and can a molecule fix it? The drug discovery process is how scientists answer that question, one careful filter at a time.
People often say “drug discovery” for the whole journey, but scientists split it in two. Drug discovery is the lab work that finds a promising molecule. Drug development is everything that follows: safety testing, clinical trials and approval. Together the journey usually takes more than 10 to 15 years and costs over $1 to 2 billion for each new approved drug.
A short history of drug discovery
For most of human history, medicines came from plants and animal products, found by trial, error and careful observation. Willow bark eased pain long before chemists made aspirin from a related compound in 1897. In the early 1900s, chemistry made synthetic drugs possible, and the modern pharmaceutical industry began. Penicillin, noticed by Alexander Fleming in 1928, opened the antibiotic era.
From the late 1970s, biology joined chemistry. Recombinant DNA technology let bacteria make human insulin, and the Human Genome Project later revealed thousands of possible drug targets. Today, computers, big data and AI help at almost every step, but the basic stages below have stayed the same.
The 6 stages of drug discovery and development
1. Target identification and validation
A target is the molecule in the body that a drug acts on. It is usually a protein, such as an enzyme or a receptor, though some targets are genes or RNA. Scientists first link the target to the disease. Then they validate it, showing that switching it up or down actually changes the disease in cells or animals.
2. Hit discovery: screening for a starting point
Next comes the search for a hit, a molecule that does something useful to the target. Robots run high-throughput screening, testing huge chemical libraries in tiny wells. Computers can also screen millions of virtual molecules by “docking” them into the target’s 3D shape. FDA notes that thousands of compounds may start as candidates, but only a small number look promising after early testing.
3. Hit to lead and lead optimisation
A hit is rarely good enough. Chemists make hundreds of small variations to turn it into a lead, then a drug candidate. They tune four things at once: potency (strong effect), selectivity (hits the target, not lookalikes), safety, and ADME, which is how the body absorbs, distributes, metabolises and excretes the drug. The form of the medicine also matters here, whether a tablet, an injection or a carrier such as lipid or polymer nanoparticles.
4. Preclinical research
Before any person takes the drug, it is tested in cells (in vitro) and in animals (in vivo). These studies ask basic safety questions: what dose is toxic, which organs are affected and how long the drug stays in the body. If the results look acceptable, the developer files an Investigational New Drug (IND) application with the FDA to begin human trials.
5. Clinical trials in people
Clinical trials run in phases. Each phase is larger and answers a harder question, and many drugs stop along the way. The figures below come from the FDA’s guide for patients.
| Phase | Who takes part | Typical length | Main question | Share moving on (approx.) |
|---|---|---|---|---|
| Phase 1 | 20–100 healthy volunteers or people with the condition | Several months | Is it safe, and at what dose? | ~70% |
| Phase 2 | Up to several hundred people with the condition | Several months to 2 years | Does it seem to work? What side effects appear? | ~33% |
| Phase 3 | 300–3,000 people with the condition | 1–4 years | Does it beat current care? Are there rarer side effects? | ~25–30% |
| Phase 4 | Several thousand people, after approval | Ongoing | Is it still safe and effective in everyday use? | Not applicable |
6. Regulatory review and post-market safety
If Phase 3 succeeds, the developer submits all of its lab, animal and human data in a New Drug Application (or a Biologics License Application for products such as antibodies). FDA review teams decide whether the benefits outweigh the risks. After approval, safety monitoring continues for as long as the medicine is sold.
Why do most drugs fail in clinical trials?
About nine out of ten drug candidates that enter clinical trials never reach patients. A widely cited review found the main reasons are a lack of clinical benefit, unmanageable side effects, poor drug-like properties and commercial or strategic decisions. This is why the early stages work so hard to pick the right target and the safest, most selective molecule.
How AI and big data are changing drug discovery
Computers have helped chemists for decades through structure-based drug design, where a protein’s 3D shape guides which molecules are made. Machine learning now adds faster virtual screening, better predictions of toxicity and ADME, and tools that predict protein structures. The FDA reports a significant increase in drug applications that include AI components. Still, AI does not skip any stage: every candidate must pass preclinical tests and clinical trials. New kinds of medicine, such as mRNA vaccines, follow the same path.
Key takeaways
- The drug discovery process runs from target to approved medicine and usually takes more than 10 to 15 years.
- The stages of drug discovery and development are target, hits, lead, preclinical, clinical trials and review.
- Clinical trials grow from about 20–100 people in Phase 1 to as many as 3,000 in Phase 3.
- About 90% of candidates that enter clinical trials fail, mostly on benefit or safety.
- AI and big data speed up the early stages, but they do not replace trials in people.
This article is for general education only and is not medical advice. Talk to a doctor or pharmacist before starting, stopping or changing any medicine, or before joining a clinical trial.
Frequently asked questions
What are the stages of drug discovery and development?
The main stages are target identification and validation, hit discovery (screening), hit-to-lead and lead optimisation, preclinical research in cells and animals, clinical trials in people (Phases 1 to 3), and regulatory review followed by post-market safety monitoring (Phase 4).
What is the difference between drug discovery and drug development?
Drug discovery is the early lab work that finds a promising molecule, from choosing a target to picking a drug candidate. Drug development takes that candidate through preclinical safety testing, clinical trials and regulatory approval.
How long does it take to discover and develop a new drug?
The whole journey usually takes more than 10 to 15 years. Clinical trials alone can take several years, with Phase 3 studies often lasting 1 to 4 years.
Why do most drug candidates fail?
About 90% of candidates that enter clinical trials fail. The main reasons are that the drug does not work well enough in people, it causes unmanageable side effects, it has poor drug-like properties, or the company stops it for commercial reasons.
How is AI used in drug discovery?
AI and machine learning help scientists screen virtual molecules, predict protein structures, forecast toxicity and how the body handles a drug, and analyse large data sets. They speed up the early stages, but every candidate still has to pass preclinical tests and clinical trials.
Sources
- U.S. Food and Drug Administration. The Drug Development Process (Steps 1–5, including Step 3: Clinical Research).
- Hughes JP, Rees S, Kalindjian SB, Philpott KL. Principles of early drug discovery. British Journal of Pharmacology, 2011. PMC3058157.
- Sun D, Gao W, Hu H, Zhou S. Why 90% of clinical drug development fails and how to improve it? Acta Pharmaceutica Sinica B, 2022. PMC9293739.
- U.S. Food and Drug Administration. Artificial Intelligence for Drug Development.