Clinical Trial Phases

Clinical drug development is conventionally divided into phases. Phase 0 (exploratory) studies, used only in some programs, give very limited exposure to a small number of people to gather early pharmacokinetic or target-engagement data. Phase 1 studies, typically in tens of healthy volunteers or patients, assess safety, tolerability and pharmacokinetics. Phase 2 studies, typically in dozens to a few hundred patients, explore efficacy and select the regimen to take forward. Phase 3 trials, typically hundreds to thousands of patients, are large randomized trials that confirm efficacy and characterize safety to support regulatory approval. Phase 4 refers to post-marketing studies and surveillance after approval, which can detect rare or long-term effects.

Tirzepatide’s obesity program illustrates late-stage testing: SURMOUNT-1 was a phase 3, double-blind, randomized trial in 2,539 adults over 72 weeks. Across all drug classes, late-stage failure is common. An analysis of 640 novel therapeutics that entered phase 3 or pivotal trials between 1998 and 2008 found that 54% failed in clinical development, most often because of inadequate efficacy (57% of failures), followed by commercial reasons (22%) and safety concerns (17%).

Each phase answers a different question, so the phase of a study tells you what it can and cannot show. A phase 1 study can show that a peptide is tolerated at the exposures tested and how the body handles it, but it is too small and short to prove benefit. A phase 2 study gives early signals of efficacy. Only large phase 3 trials, usually randomized and controlled, are designed to establish that a treatment works and to estimate how often common side effects occur.

Promising phase 2 results often fail to hold up in phase 3. Phase 2 trials are smaller, so some positive results occur partly by chance, and only the most strongly positive programs move forward. Krum and Tonkin called the resulting decline “regression to the truth”: on retesting in larger trials, effects tend to look smaller. Phase 3 also enrolls broader populations and uses harder endpoints. In one analysis, more than half of novel therapeutics entering pivotal trials failed, mostly for lack of efficacy.

Phase 4 matters because even large phase 3 programs involve thousands, not millions, of people and limited follow-up time. Rare adverse effects, interactions and long-term risks may appear only after approval, through post-marketing studies and safety reporting. For unapproved research peptides, the gap is larger: many have animal or laboratory data and, at most, small early human studies. Describing such compounds as clinically proven skips the phases that test efficacy and safety in people.

“A drug in clinical trials is almost approved”

Most investigational drugs never reach approval. Large analyses of clinical trial data have found low overall success rates, varying by disease area; one estimated a success rate of 3.4% for oncology programs. Even among drugs that reach phase 3, roughly half fail during or after pivotal trials.

“A successful phase 2 trial proves the drug works”

Phase 2 trials are designed to explore efficacy and choose a regimen, not to confirm benefit. Their smaller size makes results less certain, and effects often shrink when tested in larger phase 3 trials with broader populations. Approval normally depends on confirmatory evidence from adequately sized, controlled trials.

“Once a drug is approved, testing is over”

Approval begins a new stage of evaluation. Phase 4 studies and post-marketing surveillance monitor safety in much larger and more varied populations, can reveal rare adverse effects, and may test new uses. Large outcome trials, such as SELECT for semaglutide, can also be run to answer questions the original approval trials did not.

How many people take part in each phase?

Sizes vary widely by disease and drug, but as a general pattern phase 1 studies involve tens of participants, phase 2 studies dozens to a few hundred, and phase 3 trials hundreds to many thousands. SURMOUNT-1, a phase 3 obesity trial of tirzepatide, enrolled 2,539 adults, while the SELECT cardiovascular outcomes trial of semaglutide enrolled 17,604.

Why do so many drugs fail in phase 3?

The most common reason is that the drug does not work well enough. In an analysis of 640 novel therapeutics entering pivotal trials, 57% of failures were due to inadequate efficacy, 17% to safety concerns and 22% to commercial reasons. Earlier positive results may have been partly due to chance, smaller samples or more selected participants.

What is a phase 0 trial?

A phase 0 or exploratory study gives very limited, non-therapeutic exposure to a small number of people to check how a candidate behaves in the human body before committing to full development. It is not designed to test safety or benefit in a meaningful way and is not used in every development program.