Microdosing GLP-1s (Tirzepatide, Semaglutide): What the Evidence Shows

Short answer: "microdosing" a GLP-1 usually means taking semaglutide or tirzepatide at doses below the label's, or less often than once a week. No FDA label and no trial supports it as a regimen: the Zepbound label says its 2.5 mg starting dose "is not approved as a maintenance dosage", and Wegovy's lower steps are starting and escalation doses.^[1][2] The closest evidence is the lowest-dose arms of dose-finding trials, where smaller doses produced smaller average reductions in body weight and fewer, but still more than placebo, digestive events.^[3][4] Most microdosing depends on vials and syringes, where the FDA has documented dosing errors of 5 to 20 times the intended dose.^[5]

New to eating less? Start with what to eat when you're eating less, the guide the rest of Lighter Days builds on.

This page sets out what the term means, what the labels and trials actually say, and the risks, as of November 13, 2026. It is not a guide to doing it: we give no doses or schedules, and this is general information, not medical advice. Any change to a dose is a decision for you and your prescriber. Each medicine has its own reference page on our map of GLP-1 medications.

Microdosing GLP-1: the claims and the evidence

What is claimed	What the evidence shows
"Microdosing is an approved way to take a GLP-1."	No. No FDA label includes a microdose regimen. Zepbound's 2.5 mg and Wegovy's 0.25, 0.5 and 1 mg are starting and escalation doses, not maintenance doses.[1][2]
"Low doses work almost as well."	In trials that tested lower doses, averages fell with the dose. Semaglutide 1.0 mg weekly: −7.0% body weight at 68 weeks vs −9.6% on 2.4 mg.[6] Tirzepatide 5 mg: −15.0% at 72 weeks vs −20.9% on 15 mg.[7]
"Low doses have no digestive effects."	Fewer, not none. In tirzepatide's phase 2 trial, digestive events hit 23.1% on 1 mg vs 9.8% on placebo; in STEP 2, 57.5% on semaglutide 1.0 mg vs 34.3% on placebo.[4][6]
"There are studies on microdosing."	We found no randomized trial of a microdose regimen as it is promoted online. The data are low-dose arms of dose-finding and phase 3 trials, run at fixed doses under supervision.
"Splitting a vial is a safe way to do it."	The FDA has reports of people drawing 5 to 20 times the intended dose from multiple-dose vials, some needing hospital care.[5][8]
"Compounded or 'research' versions make it cheaper."	Compounded drugs are not FDA-approved and are not checked for quality before sale; "research use only" GLP-1s are illegal to sell for people to use.[9][8]

What do people mean by microdosing a GLP-1?

The word has no medical definition. Online, it is used for several different things: staying on a dose below the label's maintenance range, injecting a fraction of a dose more often, stretching the time between doses, or starting below the label's starting dose. It is promoted mainly for fewer digestive effects, a lower cost, or a gentler effect on appetite. Searches for "microdosing tirzepatide" alone run to thousands a month.

What these have in common is that none is a schedule the FDA has reviewed. The labels describe one way to start each medicine and a set of maintenance doses, chosen from trials that measured both benefit and harm at those doses.^[1][2]

Is microdosing GLP-1s FDA approved?

No. Here is what the labels say, quoted so you can see where the lines are, not as dosing advice:

• Zepbound (tirzepatide): treatment starts at 2.5 mg once weekly for 4 weeks, and "the 2.5 mg dosage is for treatment initiation and is not approved as a maintenance dosage." Maintenance doses are 5, 10 or 15 mg.^[1]
• Wegovy (semaglutide) injection: 0.25, 0.5 and 1 mg are the starting and escalation steps; the maintenance dose for adults is 2.4 mg (recommended) or 1.7 mg, up to 7.2 mg.^[2]
• Lower doses within the label. Both labels already tell prescribers to consider response and tolerability when choosing among the maintenance doses, such as 5 mg of Zepbound or 1.7 mg of Wegovy.^[1][2] That is a label option, set by a prescriber, not microdosing.

We found no FDA statement that addresses "microdosing" by name. Its warnings about compounded vials, dosing errors and unapproved sellers, below, cover most of how it is done.

What do trials of lower doses actually show?

Several trials tested more than one dose, which gives the best available look at what lower doses do. They were randomized, placebo-controlled and run at fixed doses, which is not how microdosing is practiced.

Trial	Who and how long	Lowest dose tested	Higher doses	Placebo
Semaglutide phase 2, dose-finding[3]	957 adults with obesity, no diabetes; 52 weeks; daily injections	0.05 mg daily: body weight −6.0%	0.1 mg −8.6%; 0.2 mg −11.6%; 0.3 mg −11.2%; 0.4 mg −13.8%	−2.3%
STEP 2[6]	1,210 adults with overweight or obesity and type 2 diabetes; 68 weeks; weekly	1.0 mg: −7.0%	2.4 mg: −9.6%	−3.4%
Tirzepatide phase 2[4]	316 adults with type 2 diabetes; 26 weeks; weekly	1 mg (the abstract reports only the range across doses)	Across 1, 5, 10 and 15 mg: −0.9 kg to −11.3 kg	−0.4 kg
SURMOUNT-1[7]	Adults with obesity or overweight, no diabetes; 72 weeks; weekly	5 mg: −15.0%	10 mg −19.5%; 15 mg −20.9%	−3.1%

How to read it, honestly:

• Lower doses did something, but not always much. The lowest semaglutide and tirzepatide arms in the longer trials beat placebo on average.^[3][6][7] In the 26-week tirzepatide trial, the smallest change across doses, −0.9 kg, was close to placebo's −0.4 kg.^[4]
• Less dose, less effect, on average. In each trial the average reduction grew with the dose.^[3][6][7]
• Digestive events also grew with the dose, but did not disappear at low doses. In tirzepatide's phase 2 trial they were 23.1% on 1 mg, 32.7% on 5 mg, 51.0% on 10 mg and 66.0% on 15 mg, against 9.8% on placebo.^[4] In STEP 2 they were 57.5% on 1.0 mg and 63.5% on 2.4 mg, against 34.3%.^[6] The semaglutide phase 2 trial reported dose-related digestive symptoms, mainly nausea.^[3]
• These are not microdose trials. The semaglutide phase 2 used daily injections of an experimental regimen, not the weekly product sold today.^[3] The tirzepatide phase 2 was in type 2 diabetes and lasted 26 weeks.^[4] None tested doses below the label's starting dose for the long term, or irregular schedules.

So the evidence answers "do lower doses work less?" (yes, on average) but not "is a self-chosen microdose regimen safe and effective?" Nobody has tested that.

What about microdosing tirzepatide specifically?

"Microdosing tirzepatide" is the most searched version. The facts are short. The lowest tirzepatide dose with long-term data for chronic weight management is 5 mg, the label's lowest maintenance dose, which produced an average −15.0% at 72 weeks in SURMOUNT-1.^[7] The 2.5 mg starting dose is not approved for maintenance.^[1] The only trial data below that come from a 1 mg arm in a 26-week diabetes trial.^[4] Zepbound comes as single-dose pens, single-dose vials and a 4-dose KwikPen;^[1] none is designed to deliver a fraction of a labeled dose. The full reference is on our tirzepatide page, and semaglutide's on the semaglutide page.

What are the risks of microdosing a GLP-1?

• Dosing errors with vials and syringes. In July 2024 the FDA reported patients drawing 5 to 20 times the intended dose of compounded semaglutide from multiple-dose vials, and clinicians miscalculating conversions between milligrams, milliliters and "units", leading to 5 to 10 times the intended dose.^[5] A poison center case series describes three people with symptoms after taking compounded semaglutide incorrectly; two had given themselves ten times the intended dose.^[10] Doing the arithmetic for a small fraction of a dose makes this kind of error easier, not harder.
• Compounded products. Most microdose plans rely on compounded vials. Compounded drugs are not FDA-approved, and the FDA does not check their safety, effectiveness or quality before sale.^[9] Since semaglutide and tirzepatide came off the shortage list, compounding copies of them is legal only in narrow cases.^[11] As of May 31, 2026, the FDA had received 990 adverse event reports for compounded semaglutide and more than 730 for compounded tirzepatide.^[8] See compounded GLP-1s explained.
• Unapproved sellers. The FDA warns against GLP-1 products sold as "for research purposes" or "not for human consumption", which may contain the wrong ingredient or the wrong amount.^[8]
• The label's warnings still apply. A smaller dose is the same drug, with the same boxed warning on thyroid C-cell tumors and the same warnings on pancreatitis, gallbladder disease and dehydration.^[1][2]
• Paying for less. A dose below the trial doses may simply do less. Approved options already include lower maintenance doses, and the makers' cash prices are lower at lower doses; our GLP-1 cost guide has the figures.

What to ask your prescriber instead

• Is a lower maintenance dose on the label, such as 5 mg of Zepbound or 1.7 mg of Wegovy, an option for me?^[1][2]
• If the digestive effects of a dose step are the problem, can the step be delayed? The Wegovy label says that if a dose is not tolerated during escalation, a prescriber can consider delaying the next step.^[2]
• If cost is the problem, which approved form and dose is cheapest under my plan or the maker's cash program?
• If I am already using a compounded product, how is my dose measured, in milligrams, milliliters or units? The FDA advises patients not to recalculate a dose themselves.^[5]

Eating on any dose

At any dose, a smaller appetite means fewer chances to get protein, fluids and micronutrients. The 2025 joint advisory on nutrition with GLP-1 therapy puts protein first, with small, frequent, nutrient-dense meals and enough fluid.^[12] Start with how much protein on a GLP-1 and, for the days after a dose step, what to eat on queasy days.

For transparency about why this page exists on our site: we make one product, Zornava Replenish, a lemon-lime daily hydration and nutrient powder for lighter appetite days, including for people on GLP-1 medications. It is a dietary supplement, not a medication, and it does not replace or imitate any medicine on this page. It is not on sale yet; early access opens February 9, 2027, and the early-access list is open now.