Three questions, one at a time
Human testing of a drug is split into phases so that each big question is answered before risking more people. We climb the ladder one rung at a time — and a drug can be stopped at any rung if the evidence turns sour.
- [[pharm-phase-i-trial|Phase I]] — a small group (often 20–100 healthy volunteers). Question: *is it safe in humans, and what does the body do to it?* Doses start tiny and rise slowly using careful dose titration.
- [[pharm-phase-ii-trial|Phase II]] — a few hundred patients who actually have the disease. Question: *does it seem to work, and at what dose?* Here we first see real signs of benefit.
- [[pharm-phase-iii-trial|Phase III]] — thousands of patients across many sites. Question: *is it better than placebo or the current standard treatment, and what are the rarer side effects?* This is the pivotal evidence regulators rely on.
Why bigger numbers come later
Each phase is larger than the last for two reasons. First, safety first: we expose few people until we have some confidence. Second, statistics: spotting a rare side effect, or proving a modest benefit is real and not luck, requires many participants. A side effect that hits 1 in 1,000 people simply cannot be seen reliably in a 50-person study.
After approval: Phase IV
Testing doesn't stop at approval. Phase IV studies follow the medicine once it's in everyday use by huge, diverse populations — people with other illnesses, on other drugs, of every age. These studies catch problems too rare to appear earlier, and the safety surveillance behind them is covered in a later guide. Throughout, fair comparison usually means a placebo control and a double-blind study design, which we explore next.