The drug development process is the organized route by which a scientific observation in a laboratory becomes a medicine a clinician can prescribe. It is best understood not as a single continuous activity but as a series of decision gates, each requiring specific evidence before the next set of investments and exposures is justified, and each involving different institutions, disciplines, and funders.
This guide follows a medicine along that route, describing what has to be demonstrated at each gate, who is responsible, and what happens after approval — because reaching authorization is not the same as reaching patients.
The Shape of the Process
Drug development is often drawn as a funnel. Very large numbers of candidate molecules enter at the discovery stage; a small fraction reach human testing; a smaller fraction still reach approval. The funnel shape is deliberate. Each stage is designed to eliminate candidates that will not succeed, as early and as cheaply as possible, before larger numbers of people are exposed.
The costly failure is the late one. A candidate that fails in a confirmatory trial has already consumed years of work and exposed substantial numbers of participants. Much of the methodological effort in modern development is aimed at moving failure earlier — better target validation, better biomarkers, better early-phase designs.
From Basic Science to a Development Programme
Basic Research
Development begins in work that was not necessarily aimed at producing a drug: studies of disease mechanisms, cell signalling, genetics, and immunology, largely conducted in universities and public research institutes and funded by agencies such as the National Institutes of Health and national research councils. This work identifies the biological targets that later programmes act on.
Translational Research
Translational research converts a mechanistic insight into something testable in humans. It includes confirming that modulating the target changes disease-relevant outcomes, developing assays and biomarkers that can measure effect, establishing suitable animal models, and producing a candidate molecule with drug-like properties.
This stage is where many findings stop. A mechanism can be real and still yield no viable medicine, because no molecule can reach the target safely, because modulating it produces intolerable effects, or because the effect in humans does not resemble the effect in models. The gap is significant enough to have its own institutional responses, including dedicated translational research centres and public-private partnerships.
Candidate Selection
A formal decision gate. The organization commits to a specific molecule, having assessed potency, selectivity, pharmacokinetic properties, early safety signals, manufacturability, and the competitive and clinical landscape. Everything after this point is built around that molecule.
Preclinical Development: Establishing the Case for Human Testing
Preclinical development assembles the evidence a regulator requires before humans are exposed. It comprises pharmacology establishing the intended effect and dose-response, toxicology conducted under Good Laboratory Practice in more than one species, pharmacokinetic characterization, and manufacturing of clinical-grade material of defined quality.
The output is an application to begin clinical testing — an Investigational New Drug application in the United States, a clinical trial authorisation in European Union member states — reviewed alongside independent ethics committee approval of the protocol and consent materials.
Clinical Development: Three Questions in Sequence
Clinical development answers three questions in order, and the phase structure exists to keep them separate.
Can it be given to humans safely, and at what dose?
Early-phase studies enrol small numbers, typically healthy volunteers except where the candidate is expected to be toxic. They characterize tolerability, dose range, and how the body processes the compound. They are not designed to show benefit.
Does it do anything in the intended disease?
Mid-phase studies enrol patients and look for a signal of biological or clinical effect, refine dosing, and expand safety observation. This is the gate at which the largest number of programmes terminate, and terminating here is the system working as intended.
Does it produce meaningful benefit compared with the alternatives?
Confirmatory trials are randomized, controlled, usually blinded, and large enough to detect differences that matter clinically and adverse effects too uncommon to appear earlier. They are typically run across many centres and countries, frequently with contract research organizations managing site operations, monitoring, and data handling on the sponsor's behalf.
Who Does What
- Universities and public research institutes generate much of the underlying biology and many original targets.
- Biotechnology companies often carry candidates from discovery through early clinical testing, frequently founded around a specific mechanism or platform.
- Pharmaceutical companies typically lead late-stage clinical development, regulatory submission, manufacturing at scale, and global distribution, often acquiring or licensing assets from smaller firms.
- Contract research and manufacturing organizations provide specialist trial operations, laboratory services, and production capacity.
- Regulators authorize trials, review marketing applications, inspect facilities, and oversee post-market safety.
- Investigators, clinical sites, and participants conduct and take part in the trials on which everything depends.
- Patient organizations increasingly contribute to trial design, endpoint selection, and recruitment.
Regulatory Submission and Review
A marketing application assembles clinical efficacy and safety data, non-clinical pharmacology and toxicology, chemistry and manufacturing information, and proposed labelling. Regulators assess whether the benefit-risk balance is favourable for a defined indication and population, whether the product can be manufactured consistently, and whether the labelling represents the evidence accurately.
Review is iterative: agencies pose questions, request analyses, may convene expert advisory committees, and inspect trial and manufacturing sites. Possible outcomes include approval, approval with a narrower indication or additional risk-management obligations, a request for further data, or refusal. Expedited pathways exist for serious conditions with unmet need and alter the sequence and timing of evidence rather than the underlying standard.
After Approval: Reaching Patients
Authorization permits marketing. It does not by itself mean patients receive the medicine. Several further steps usually intervene.
- Health technology assessment and reimbursement. Many health systems assess clinical and economic value to decide whether and on what terms a medicine will be funded. A product approved by a regulator may be reimbursed in one country and not another, or restricted to a subgroup.
- Clinical guidelines. Professional bodies appraise the evidence and position the new option relative to existing treatments, strongly influencing prescribing.
- Supply and distribution. Manufacturing capacity, cold chain requirements, and distribution networks determine practical availability, and shortages can arise from any point in the chain.
- Clinician and system readiness. Some products require diagnostic testing, specialist administration, or monitoring infrastructure that must be in place before use can spread.
Global access adds a further layer. Regulatory approval in one region does not extend to others, and mechanisms such as WHO prequalification, regional regulatory reliance, licensing arrangements, and tiered pricing shape how quickly a medicine becomes available in lower-income settings.
Continuing Evidence Generation
Once in use, a medicine is studied under conditions no trial can replicate: unselected patients, long durations, concurrent illnesses and medications. Pharmacovigilance systems collect adverse event reports and assess signals. Post-authorization studies, sometimes mandated by regulators, examine long-term safety and effectiveness. Analyses of registries and routinely collected health data provide real-world evidence complementing trial findings.
This continuing surveillance can change a product's status at any point in its life, through label updates, added warnings, restricted indications, or withdrawal.
Sources
- U.S. Food and Drug Administration — drug development and review process; real-world evidence programme
- European Medicines Agency — from laboratory to patient; post-authorisation obligations
- U.S. National Institutes of Health — translational research programmes
- World Health Organization — prequalification of medicines; access to essential medicines
- International Council for Harmonisation — efficacy, safety, and quality guidelines
- ClinicalTrials.gov — study registration and results
