Microdosing is the GLP-1 world's fastest-growing improvisation: staying on starter doses, stretching pens, halving vials, chasing fewer side effects or lower bills. Some of it is rational medicine done informally; some of it is arithmetic roulette. This guide separates the two, label-first.
What the labels actually say
Zepbound and Mounjaro position 2.5 mg explicitly as an initiation dose, not maintenance: the labels escalate to 5 mg and beyond, and the trials that produced the headline results ran at 10 to 15 mg. Staying at 2.5 mg indefinitely is off-label improvisation, though a clinician can legitimately choose it. Wegovy follows the same pattern with its own steps. Nothing in any label describes 'microdosing'; the term lives entirely outside the approved universe.
The three honest reasons people microdose, graded
- Side-effect management: the most defensible version. Slower escalation and lower plateaus genuinely reduce GI effects, and clinicians do this formally; the trials themselves used gradual titration for exactly this reason.
- Cost stretching: splitting doses to make pens or vials last longer. Understandable economics, but efficacy at fractional doses is not established: the appetite effect is dose-dependent, and half a dose can mean most of the cost with a fraction of the result.
- Maintenance after goal weight: an emerging, plausible use (keep some signal, reduce burden), with early clinical practice but no dedicated trial defining the right maintenance floor yet.
Where the arithmetic gets dangerous
Single-dose pens are engineered to deliver everything they hold; extracting partial doses from them means operating outside the device's design, with real dosing-precision risk. Compounded vials are where splitting actually happens, and they are also where the FDA has documented 5-to-20-fold overdose errors from mg/ml/unit confusion. If a vial and syringe are involved, the conversion math is the safety layer: our calculator exists for precisely this case, and its flags catch the classic half-dose errors (entering mg for mcg, draws too small to measure).
The evidence status, plainly
No published trial validates microdosing protocols for weight management: not for efficacy floors, not for maintenance dosing, not for side-effect-free thresholds. What exists is the dose-response data from the main trials (lower dose, less effect, fewer GI events) and accumulating informal clinical practice. That makes microdosing a clinician-judgment zone, not an evidence-backed protocol, and anyone selling a 'microdose plan' as established science is ahead of the literature.
Frequently asked questions
Can you stay on 2.5 mg of Zepbound?
The label treats 2.5 mg as a four-week initiation step, so staying there is off-label, but a prescriber can choose it deliberately, typically for tolerability. Expect meaningfully less weight loss than the 10 to 15 mg trial results: dose-response is real in both directions.
Does microdosing GLP-1s work?
For side-effect management, slower-lower dosing demonstrably helps. For weight loss at fractional doses, no trial defines what still works; the main trials show effect shrinking with dose. It is unstudied territory navigated by clinician judgment.
Can you split a Zepbound pen?
The single-dose pens are not designed for partial delivery, and improvising around that design is a precision risk. Dose-splitting in practice happens with compounded vials, where correct mg-to-units math becomes the entire safety margin.