Before you read. This page describes what the FDA label and its trial programme say about timing. It is educational, it is not a prediction for you, and it cannot tell you whether your treatment is working — that judgement needs your numbers and your prescriber. Nothing here is a reason to change a dose, skip a step, or climb the ladder faster than you were told to.
Quick answer
Three dates, all of them from the label.
Day 30. The starting dose is 3 mg once daily for the first 30 days, and the label describes it in plain language as an initiation dose that is not effective for glycemic control. It is there to let the gut adjust — nothing more. If nothing has happened by week three, that is the design working, not the drug failing.
Around day 60. Day 31 is when the first therapeutic dose starts — 7 mg — and steady-state exposure is reached after a further four to five weeks of daily dosing. So the earliest point at which you are both on a working dose and at a stable level is roughly two months in.
Week 26. That is where the trials put their primary measurement. Not because nothing happens before it, but because it is where the effect on the endpoint the drug is licensed for was judged.
Everything else on this page is those three dates explained. If you are past all of them and still asking, that is a different question, and six things to check before it is the drug is where it goes.
Days 1 to 30 — the month that is not a treatment month
This is the single most useful sentence in the dosage section, and almost nobody is shown it:
Day 1 to 30: Recommended starting dosage is 3 mg orally once daily for 30 days (this dosage is not effective for glycemic control).
A parenthesis is doing a lot of work there. The 3 mg tablet is not a small dose of a working treatment. It is a tolerance step — the label's stated reason for stepping the dose at all is to reduce the risk of gastrointestinal adverse reactions, and 3 mg is the first rung.
Which reframes the first month entirely. People spend it watching for an effect that the label does not claim will be there, and often conclude the drug does not work for them before they have taken a dose intended to do anything.
What the first month is for is finding out how your gut handles it. The label reports gastrointestinal reactions clustering during dose escalation, and what it reports and when goes through the numbers. If month one is uneventful, that is useful information about month two.
Day 31 onward — the first dose meant to do something
On day 31 the dose goes to 7 mg once daily. On day 61, if additional glycemic control is needed, it can go to 14 mg.
That is the whole ladder — three strengths, thirty days each. It is worth saying explicitly that 14 mg is not the destination — the label frames it as conditional on whether more control is needed, and a stable 7 mg is a complete treatment for the people it suits.
The thirty days is not padding. It is long enough for the previous step to reach a stable level before another step is added on top, which is the point of the next section.
What steady state means when the half-life is a week
Here is where the arithmetic of this drug diverges from most tablets.
Semaglutide's elimination half-life is approximately one week. A drug reaches steady state — the level where each day's dose replaces what was cleared — after roughly four to five half-lives. The label states it directly: steady-state exposure was achieved following 4 to 5 weeks of semaglutide tablets oral administration.
So each rung of the ladder takes about a month to become what it is going to be. Start 7 mg on day 31 and the exposure at that dose is still climbing through most of February. The 30-day hold before the next step exists so that the increase happens from a stable place rather than a moving one.
The label also gives the levels themselves: mean steady-state concentration of 6.7 nmol/L on 7 mg and 14.6 nmol/L on 14 mg, roughly doubling with a doubling of the dose. Semaglutide exposure increases in a dose-proportional manner, which is the property that makes the ladder predictable.
| What is happening | |
|---|---|
| Days 1–30 | 3 mg. Initiation dose, not effective for glycemic control |
| Days 31–60 | 7 mg begins. Exposure climbing toward steady state through most of this month |
| Around day 60 | First point at which you are on a therapeutic dose and at a stable level |
| Day 61 onward | 14 mg if additional glycemic control is needed. Another 4–5 weeks to its own steady state |
| Week 26 | Where the trials measured the primary endpoint |
None of that is a prediction about you. It is the schedule the drug is on regardless of who is taking it.
Week 26, and what the trials were actually measuring
The Rybelsus trial programme put its primary endpoint at week 26 — the change in HbA1c, in adults with type 2 diabetes. That is the measurement the drug is licensed on, and the broader pooled safety population was treated for a mean of about 60 weeks.
Two things follow from that, and both are easy to miss.
HbA1c is a slow measurement by construction. It reflects average glucose over roughly the preceding three months, so it cannot show you a change from last week. Even if the drug started working on day 31, a test taken in week 8 is still averaging in weeks 1 to 4, when by the label's own account not much was meant to be happening.
The endpoint is glycemic. Rybelsus is indicated as an adjunct to diet and exercise to improve glycemic control in adults with type 2 diabetes, and to reduce the risk of major adverse cardiovascular events in adults with type 2 diabetes at high risk. Those are the two things the label claims. Judging the timeline by a different measurement is judging it against something it was not licensed to do.
"Has anything changed?" is a question about dates, and dates are what memory loses first. Takely puts the dose changes and the daily notes on one timeline, so week 9 can actually be compared with week 4. Open Takely in your browser →
What people notice first, and it is not the number
Before any lab result moves, the commonly reported early change is appetite.
The label lists decreased appetite as an adverse reaction — reported by 9% of patients on 14 mg and 6% on 7 mg, against 1% on placebo. It is filed under adverse reactions because that is where a label puts anything that happens to more people on the drug than off it, but it is also a description of the mechanism working.
The same slowed gastric emptying that produces the drug's effect on post-meal glucose is felt as fullness arriving sooner and lasting longer. For a lot of people that is the first perceptible sign that something has changed, and it tends to arrive during the second month rather than the first. What has evidence for quieting food noise covers the related experience and what is actually known about it.
A caution about reading effects backwards. Feeling nothing is not evidence that nothing is happening — and feeling something is not evidence that the endpoint is moving. Glycemic control is measured, not sensed. The reason to track how you feel is that it gives your prescriber a second axis to read the numbers against — not that it substitutes for them.
The one thing that changes the timeline
Everything above assumes the dose is being absorbed, and with this tablet that assumption is doing real work.
Absolute bioavailability is 0.4% to 1%. Absorption happens in the stomach through a narrow window, and the label fences it: empty stomach, morning, no more than 4 ounces of plain water, nothing else to drink, swallowed whole, then at least 30 minutes before food, other drinks or other oral medicines.
A routine that breaks the window most mornings is not a slower version of the same timeline. It is a lower dose than the one on the box. Why so little of each tablet gets in explains the mechanism, and the 30-minute wait, step by step covers where mornings actually break.
This is the first thing worth checking before concluding that two months have passed without effect, because it is the only variable on this page you control.
What to measure, and against what
Four things, and the fourth is the one people skip.
The date of each dose change. Every timeline on this page runs from a dose change, not from the day you started. Reaching 7 mg late because a step was held makes week 12 look different from the schedule above, and that is worth knowing rather than puzzling over.
Appetite, as a daily line. It moves before anything measurable and it is the earliest signal you have access to.
Symptoms with dates. The label ties gastrointestinal reactions to escalation. Symptoms that appear four days after an increase and settle within a fortnight tell a different story from symptoms that persist on a stable dose.
Whether the morning was clean. Window kept or window broken, one bit per day. Over eight weeks that column is the difference between "the drug is not working" and "I have been taking a fraction of it," and nobody can reconstruct it afterwards. Counting the days from the dose change rather than from the start is what makes the other three legible.
When the calendar has run out of excuses
If you are past day 60 on a therapeutic dose, your mornings have been clean, and the measurements your prescriber cares about have not moved, then the timeline is no longer the explanation.
That is a real situation with real causes, and it deserves its own list rather than a paragraph here. Six things to check before it is the drug works through them, in the order worth checking.
The one thing not to do is climb the ladder privately. The 30-day holds and the conditional wording on 14 mg are both there for reasons the label states, and taking more than one tablet a day is ruled out explicitly.
Frequently asked questions
How long does Rybelsus take to work?
The label's schedule puts the first therapeutic dose at day 31, with steady state reached after a further four to five weeks — so roughly two months before you are on a working dose at a stable level. The trials measured their primary endpoint at week 26.
Why does the 3 mg dose not seem to do anything?
Because the label says it does not. The dosage section describes 3 mg as the starting dosage for the first 30 days and states in parentheses that this dosage is not effective for glycemic control. It is a tolerance step intended to reduce gastrointestinal reactions before a therapeutic dose starts.
How long before I notice a change in appetite?
The label lists decreased appetite as an adverse reaction at 9% on 14 mg and 6% on 7 mg against 1% on placebo, but gives no timing for it. In practice it is the earliest thing most people report, and it follows the dose rather than the calendar — which usually means the second month rather than the first.
Should my dose be increased if nothing has happened after a month?
That is a prescriber's decision, and the label's own schedule already increases the dose on day 31. What matters is that the increase happens on the ladder rather than ahead of it: the 30-day holds exist to reduce gastrointestinal reactions, and the label caps intake at one tablet per day.
Does the tablet work more slowly than the injection?
The half-life is about a week either way, and steady state takes four to five weeks either way. What differs is that the tablet's dose is absorbed through a narrow stomach window at under 1% bioavailability, so a broken morning routine costs exposure in a way an injection cannot.
When will my HbA1c show a change?
HbA1c reflects average glucose over roughly the previous three months, so a test in the early weeks is still averaging in the initiation month. Your prescriber decides when to test; the trials chose week 26 for the primary measurement.
Primary sources and further reading
- FDA Prescribing Information — RYBELSUS (semaglutide) tablets and OZEMPIC (semaglutide) tablets, sections 1, 2.1, 2.2, 6.1, 12.3 and 14 ↗
- FDA label via DailyMed — the current RYBELSUS listing and revision history ↗
Label revisions change dosing text and trial descriptions. Everything above was read off the prescribing information on 8 September 2026.