Retatrutide safety: what a phase 2 trial can and cannot tell you yet

Key takeaways

The human safety evidence is one randomized phase 2 trial: 338 adults, 48 weeks.
That size can find common side effects. It structurally cannot find rare ones.
Phase 3 is where durability and rare harms get characterised, and it hasn't reported.
Separately, our lab data shows retatrutide vials missing their labelled dose by a median of 11.7%, up to 43.5%.

The honest answer to "is it safe"

It can't be answered yet, and that isn't a dodge. It's the actual state of the evidence.

No regulator has authorized retatrutide for any use.¹ That matters more than people realise, because approval is the point at which an authority formally judges that a compound's benefits outweigh its risks for a defined population and purpose.

Nobody has made that judgement here. Not because it was made and came out badly, but because the studies it depends on haven't finished.

What exists is a randomized phase 2 trial in 338 adults over 48 weeks.

That's a real trial and a real result. It's also a specific size and a specific duration, and those numbers determine exactly what it can and can't tell you.

What a 338-person, 48-week trial can find

Quite a lot, actually. It's worth being fair about this rather than dismissive.

A study of that size can characterise common side effects, the ones that turn up in a meaningful share of participants. It can compare rates against a control group. It can measure how many people stopped taking it and why, which is often the most honest safety signal a trial produces.

It can also detect problems severe enough to stop a trial. None of that is nothing.

What it structurally cannot find

Here's the part that gets skipped, and it's arithmetic rather than opinion.

A harm affecting one person in a thousand will usually not appear at all in a study of 338 people.

You can't detect what you're statistically unlikely to observe. That isn't a criticism of the trial design; mid-stage trials aren't sized for rare events, and nobody claims otherwise.

Question	Can phase 2 answer it?
Common side effects in enrolled participants	Yes
Dropout rates and reasons	Yes
Rare harms	No
Effects beyond 48 weeks	No
Behaviour in populations not enrolled	No
Interactions across long-term use	No

Phase 3 exists to answer the bottom half of that table.² Larger populations, longer follow-up, broader enrolment. Our profile states that phase 3 development is intended to determine whether the benefits, risks, and durability hold in larger populations.

Intended to determine. Not confirmed.

And a clean phase 2 doesn't forecast a clean phase 3. Programmes do surface problems at that stage, which is the reason the stage is mandatory. We work through this in why evidence maturity beats online momentum.

The safety question nobody frames as one

Now a factor specific to buying an unapproved compound, and it comes from our own data rather than any trial.

Our testing ledger holds 44 published lab records for retatrutide, more than for any other compound we track.³ Of the 40 carrying dose data:

	Result
Identity confirmed	Every record
Purity range	99.07% to 99.94%
Median absolute miss from label	11.7%
Vials 10% or more over label	24 of 40
Widest single miss	+43.5%
Vials within 5% of label	7 of 40

Only 7 of 40 vials landed within 5% of what their label claimed.

The material is clean and correctly identified. The quantity is not reliable, and one vial sold as 10mg measured 14.35mg.

A trial's safety findings are attached to known quantities administered under supervision. Neither condition holds when the labelled amount is wrong by a double-digit percentage and there's no monitoring involved.

That's not a claim about what retatrutide does to anyone. It's a statement about how far trial evidence can travel, and the answer is that it doesn't travel to a vial whose contents differ from its label.

For how to read these documents yourself, see our guide to certificates of analysis.

What would change this picture

Worth naming, so the article has a shelf life.

Phase 3 reporting, with durability and rare-harm data across a larger population.

A regulatory decision, in either direction, which would be the first formal benefit-risk judgement on this compound.

More supply data, particularly whether dose variance narrows as the market matures.

We update our profile as primary sources change, and we bump the date on these articles when the figures move.

Where this leaves you

Promising is accurate. Safe is not established. Those coexist, and holding both is the whole discipline here.

The phase 2 result is genuinely more than most compounds discussed as peptides will ever have. It is also mid-stage evidence about common effects in 338 people over 48 weeks, which is a narrower claim than it gets used for.

Anyone telling you retatrutide's safety is established is describing a study that hasn't reported yet.

Background on the compound itself: what retatrutide is.

Frequently asked questions

Is retatrutide safe?

That question cannot be answered from the evidence that exists. A randomized phase 2 trial in 338 adults over 48 weeks can characterise common side effects in the people it enrolled. It cannot establish rare harms, effects beyond 48 weeks, or behaviour in populations it did not study. No regulator has authorized the compound, which means no authority has judged its benefits to outweigh its risks for any use.

What side effects were seen in the retatrutide trial?

We do not publish trial side-effect tables here, and we would point you to the primary trial publication and our profile's graded outcomes for that detail. What matters structurally is that any side-effect profile from a 338-person study describes common effects, and that rare events are not detectable at that sample size.

Why does trial size matter for safety?

Because a harm that occurs in one person per thousand is unlikely to appear at all in a study of a few hundred. Detecting rare events requires large populations and long follow-up, which is a core reason phase 3 programmes exist and why approval requires them.

Does a clean phase 2 mean phase 3 will be clean?

No. Phase 3 programmes sometimes surface harms or durability problems that mid-stage trials did not, which is precisely why they are required. A phase 2 result is a reason to continue investigating rather than a preview of the conclusion.