Retatrutide Dosage: What Each Trial Dose Did, From 1 mg to 12 mg

What retatrutide dosage did the trials use?

Retatrutide (Eli Lilly code LY3437943) is a single peptide that activates the GIP, GLP-1 and glucagon receptors. Its human dose range was mapped in three steps: a 12-week phase 1b multiple-ascending-dose study, two 36 to 48-week phase 2 trials, and a phase 3 program that began in 2023. The table lists every weekly maintenance dose that has been published with results, and the starting dose that led to it.

Two things stand out. First, no trial has gone above 12 mg a week, so 12 mg is the ceiling of the human evidence. Second, the phase 3 designers dropped 1 mg and 8 mg and inserted 9 mg, a choice the Lancet authors say was informed by the phase 2 dose-response data.

How much weight loss did each retatrutide dose produce?

The clearest dose-response data comes from the NEJM phase 2 obesity trial. Participants had a BMI of 30 or higher, or 27 to 30 with a weight-related condition, and were followed for 48 weeks. The chart shows the least-squares mean percentage change in body weight at the primary endpoint (24 weeks) and at the end of treatment (48 weeks).

The curve flattens at the top. Going from 1 mg to 4 mg roughly doubled the effect, 4 mg to 8 mg added another 5.7 points, and 8 mg to 12 mg added 1.4 points. Weight was still falling at week 48 in the higher dose groups, which is one reason the phase 3 trials run to week 80.

How was the retatrutide dosage escalated?

Every maintenance dose above 1 mg was reached by steps, not in one jump. In the phase 2 obesity trial the registry describes the 12 mg arm as "2 mg starting dose followed by 4 mg, 8 mg and then 12 mg", with escalation "up to Week 12". That is three steps in twelve weeks, and the same 2 to 4 to 8 pattern was used for the 8 mg arm. Two arms were deliberately started at 4 mg instead of 2 mg to test whether the gentler start mattered; it did, for gastrointestinal tolerability, though not for the final weight result.

The phase 3 TRIUMPH trials use a "fixed dose escalation regimen" that takes 16 weeks, followed by 64 weeks of maintenance. The design paper does not print the intermediate steps, so this site does not guess them. Both programs allow a permanent dose reduction when gastrointestinal events or poor oral intake do not settle. The full week-by-week layout is in the dosing schedule guide, and the tolerability findings in the titration guide.

Which retatrutide dose did the most?

By the numbers, 12 mg. It produced the largest mean weight change and the largest share of participants crossing each threshold: at 48 weeks, 100% lost at least 5%, 93% lost at least 10% and 83% lost at least 15% of their body weight. The 8 mg group was close behind at 100%, 91% and 75%.

Whether the extra 1.4 points from 8 to 12 mg is worth the dose-related increase in gastrointestinal events is exactly the question the phase 3 program was built to answer with larger numbers, which is why 9 mg sits between them in TRIUMPH. In the liver substudy, 8 mg and 12 mg were almost identical (liver fat down 81.4% and 82.4% at 24 weeks), and in the type 2 diabetes trial the 8 mg and 12 mg HbA1c results were within 0.03 points of each other. All of those figures, by dose and by trial, are collected in the results page.

Is there a standard retatrutide dosage?

No. Retatrutide has no marketing authorisation, so there is no label and no approved dose. The closest thing to a consensus is the phase 3 choice of 4, 9 and 12 mg once weekly, because that is what Lilly and its investigators considered worth testing in more than 5,800 people. The first phase 3 publication, the 40-week TRANSCEND-T2D-1 trial in the Lancet in June 2026, reported mean weight changes of -11.5% at 4 mg, -13.9% at 9 mg and -15.3% at 12 mg in adults with type 2 diabetes, alongside HbA1c reductions of 1.69 to 1.94 percentage points.

Any text that presents a "protocol" outside these ranges, or a dose adjusted to body weight, is not drawn from the trials: none of the published studies dosed by weight, and none went below 0.5 mg or above 12 mg.

How does the dose relate to side effects?

The NEJM paper describes the adverse events in one sentence worth quoting in substance: the most common events were gastrointestinal, they were dose-related, mostly mild to moderate, and partially mitigated by the 2 mg starting dose compared with 4 mg. Heart rate increased in a dose-dependent way, peaked at 24 weeks and declined afterwards. In the type 2 diabetes trial, mild-to-moderate gastrointestinal adverse events were reported by 13% of the 0.5 mg group and by 50% of the 24 participants in the 8 mg fast-escalation group, against 35% across all retatrutide groups combined and 13% on placebo, which is the strongest single argument in the data for starting low and climbing slowly.

The side-effect profile by dose, including the heart-rate data, is covered in more depth on the sister site https://retatrutiderisk.com, in particular its side-effects article.

Retatrutide dosage in type 2 diabetes

The diabetes trials used the same ladder with one extra rung at the bottom: 0.5 mg. In the 2023 Lancet phase 2 trial, HbA1c at 24 weeks fell by 0.43 points at 0.5 mg, 1.30 to 1.39 at 4 mg, 1.88 to 1.99 at 8 mg and 2.02 at 12 mg, versus 0.01 with placebo and 1.41 with dulaglutide 1.5 mg. Body weight at 36 weeks fell 3.19% at 0.5 mg and 16.94% at 12 mg. The dose-by-dose detail, including the 2026 phase 3 numbers, is in the type 2 diabetes dose guide.

What a research vial label means versus a trial dose

Research suppliers sell retatrutide as lyophilised powder in vials labelled 5 mg, 10 mg or 20 mg. That number is the total peptide in the vial, not a weekly dose. A 10 mg vial reconstituted with 2 mL of diluent gives 5 mg/mL, so 4 mg (a phase 3 dose) occupies 0.8 mL and 12 mg would need more than one vial. The arithmetic, with a calculator, is in the reconstitution guide, and the mapping of vial sizes to trial doses in the vial-size guide. Suppliers who publish a certificate of analysis per batch make it possible to check that the label matches the contents; one such listing is at See it at OXpeptides (research use only).