| Phase | What a positive result can show | Planning use |
|---|---|---|
| Phase 1 trial | Safety and dosage in a small group, often healthy people | Treat it as a safety step. It is not evidence that the treatment helps patients. |
| Phase 2 trial | Preliminary data on whether the drug works | Read it as an early signal that still needs confirming in a larger trial. |
| Phase 3 trial | How the new treatment compares with standard treatment or a placebo | The closest thing to a head-to-head answer before approval. Check the comparator and the outcome measured. |
Phase descriptions in the table follow the UC San Diego Extended Studies guide to the four phases and the EOS Network summary of clinical trial phases.
Clinical trial phases explained: Phase 1 asks about safety and dosage, Phase 2 about early signs of benefit, Phase 3 about comparison with standard care, and Phase 4 about an approved medicine. The phase caps what a headline can claim.
Readers who follow medical news today meet these labels constantly. A headline that says “promising” or “effective” often leaves out which phase produced the result.
Reading a headline by its phase
A phase label tells a news reader which question a trial was designed to answer. The NCATS glossary of clinical trial phases says the phase is based on the study’s objective, the number of participants and other characteristics, so each label carries a different weight.
- Phase 1 headline. Read it as safety and dose news. The EOS Network describes a small group, often healthy people, tested for safety and dosage. Such a trial is not built to show that a treatment helps people with the disease.
- Phase 2 headline. Treat a positive result as a reason for a bigger trial. UC San Diego Extended Studies describes Phase 2 as gathering preliminary data on whether the drug works. “Preliminary” is the operative word, and it argues against wording such as “proven” or “cure”.
- Phase 3 headline. This is the first place a headline can fairly describe a benefit against something. Ask what the comparator was, whether standard treatment or a placebo, and which outcome was measured. Then check whether the reported gain is relative or absolute, as the guide on relative risk vs absolute risk explains.
- Phase 4 headline. Read it as monitoring of a medicine that is already approved. Covenant Health describes these as observational or registry studies that can find side effects not discovered in earlier phases. A new safety finding here is a result of that monitoring, not a sign the trials were run badly.
When a story gives no phase at all, look up the trial’s record before sharing it (the registry section below shows how). For everything beyond the phase, including sample, design and endpoints, the walkthrough on how to read a medical study covers the next checks.
How each phase works
The phases run in order, with the question narrowing at each step and the group of participants growing. Sizes and durations below are ranges from published summaries, and they differ between sources, so treat them as ballparks.
Before Phase 1: the IND
No human trial starts until a regulator has seen the plan. Covenant Health notes that the U.S. Food and Drug Administration must review and approve a specific Investigational New Drug application, or IND, before trials in humans. A review of first-in-human studies in PubMed Central says these studies are always preceded by regulatory filings to health authorities, such as US IND applications. A headline about a drug “entering human trials” therefore describes a filing that has already cleared review, not evidence of benefit.
Phase 1: safety and dose
First-in-human trials are conducted primarily to determine the safe dose range for further clinical development, according to the PubMed Central review. UC San Diego Extended Studies puts a Phase 1 group at 20-80 healthy volunteers and the study length at several months to a year. Covenant Health gives a wider figure of about 20-100 participants. Either way, the group is too small to say much about effectiveness.
Phase 2: a first look at whether it works
Phase 2 moves to people with the condition and asks for preliminary evidence of benefit. The EOS Network says it typically includes 100 to 300 participants. A positive Phase 2 result means the signal justified continuing. It does not mean the drug has beaten standard care.
Phase 3: comparison at scale
Phase 3 is where the comparison group arrives. The EOS Network describes comparing the new treatment with standard treatment or a placebo, with hundreds to 3000 participants and a length of three to five years. Covenant Health gives a range of around 300-3,000 participants lasting 1-4 years. Because these trials run for years, a Phase 3 readout usually lands long after the Phase 2 news that preceded it.
Phase 4: after approval
Phase 4 trials, in Covenant Health’s description, are observational or registry studies of a drug that is already on the market. They can identify side effects that were not discovered in earlier phases. The EOS Network says this stage is much longer than the previous one and can take 10-20 years after licensing. Long follow-up of this kind is why Phase 4 safety updates can appear years after a medicine reaches pharmacies.
How many candidates advance
Most candidates do not clear every phase, which is why an early positive headline rarely predicts an approval on its own. UC San Diego Extended Studies says approximately 70% of drugs that enter clinical trials pass Phase 1 and approximately 33% of treatments that begin Phase 2 move to the next phase. Covenant Health, citing the FDA, gives about 70% for Phase 1 and approximately 25-30% of Phase 3 drugs moving to the next phase or approval.
The Phase 2 and Phase 3 shares are both lower than the Phase 1 share. Read the figures as rough guides to attrition at each step rather than something to combine into a single odds-of-approval number.
Reading the phase on a ClinicalTrials.gov record
The phase field on a ClinicalTrials.gov entry is the quickest way to check what kind of trial a headline describes. The labels follow the NCATS glossary, and a few habits make them easier to use.
- Do not read “Not Applicable” as a failure. NCATS says the label is used for trials without FDA-defined phases, including trials of devices or behavioral interventions.
- Remember what the label reflects. NCATS says phase is based on objective, number of participants and other characteristics. It describes the trial’s design and goal, not whether the result was good.
Frequently asked questions
Is a Phase 0 trial required?
No. Covenant Health notes that some Phase 0 trials are conducted in humans, but they are not required, so many drugs go straight from preclinical work to Phase 1.
Does a Phase 1 trial only involve healthy people?
Not always. The EOS Network describes Phase 1 as testing an experimental treatment on a small group of often healthy people, which leaves room for other participant groups.
Do the phases have fixed lengths?
No. Published ranges differ. A Phase 3 trial is described as lasting three to five years by the EOS Network and 1-4 years by Covenant Health, so quote a range rather than a single figure.
