How Long Should You Give a New Antidepressant?
By Lance Demaline • July 31, 2026
Quick Answer: Longer than most people do, but not indefinitely. Antidepressants need several weeks at a therapeutic dose before you can fairly judge whether they're working, and side effects usually show up before benefits do. Your prescriber will have a specific review point in mind. What matters is that the trial is long enough, at a high enough dose, taken consistently, because those three things are what make it count as a real trial rather than an inconclusive one.
Most people either stop too early or wait far too long. Stopping at two weeks because nothing has changed means you never find out whether the medication would have worked. Staying on something for two years because it's "probably doing something" costs you the same time in a different way. This post covers what should happen in the early weeks, what makes a trial count, and why keeping a record of it matters more than you'd expect.
Why it takes weeks rather than days
Antidepressants change neurotransmitter levels fairly quickly, but the downstream changes that actually shift mood take longer to develop. That gap is why the first stretch on a new medication can feel like the worst of both worlds: you're getting the side effects without any of the benefit.
This isn't the medication failing. It's the normal shape of the thing. Knowing that in advance is the single best predictor of whether someone sticks with a trial long enough to learn anything from it.
Improvement is also usually gradual rather than sudden. People often notice it first in sleep, appetite, or energy, and only later in mood. Sometimes the people around you notice before you do.
What should happen in the first few weeks?
Expect side effects early. Nausea, headaches, restlessness, changes to sleep, and digestive upset are common at the start and often ease as your body adjusts. Your prescriber can tell you which ones are expected to settle and which ones aren't.
When to call sooner rather than waiting it out:
- Your mood gets noticeably worse rather than better
- New or increasing agitation, restlessness, or inability to sit still
- New or worsening thoughts of harming yourself
- Any side effect that's severe, or that you can't function around
- A rash, or anything that feels like a physical reaction rather than an adjustment
None of these are reasons to be stoic. Early worsening is uncommon but it's real, and it's specifically why prescribers want to see you sooner after a start or a dose change rather than later.
If you're having thoughts of harming yourself, call or text 988, or go to your nearest emergency department. Don't wait for your next appointment.
Don't stop a medication abruptly on your own either. Several antidepressants cause discontinuation symptoms if stopped suddenly, and those get mistaken for relapse. If you want off it, tell your prescriber that and they'll taper you.
What actually counts as an adequate trial
Three things have to be true before a medication trial tells you anything useful:
A therapeutic dose. Many people are titrated up gradually and then stay at a starting dose that was never meant to be the destination. A trial at the lowest dose isn't a failed trial. It's an incomplete one.
Enough time at that dose. The clock starts when you reach the therapeutic dose, not when you first took the tablet. This is where a lot of "failed" trials fall apart on review.
Consistent use. Missed doses, stopping and restarting, or gaps caused by cost or pharmacy issues all muddy the result. Say so honestly at your review appointment. Prescribers are not shocked by this, and it changes what they recommend.
If any of the three is missing, the honest answer is that you don't yet know whether the medication works for you.
Still weighing your options?
A quick consultation answers more than another hour of reading.

What happens if it isn't working
Your prescriber has a few directions from here, and which one they pick depends on whether you got a partial response or nothing at all:
- Optimize the dose. Partial response at a mid-range dose often means there's room to go up.
- Switch. No response at a full dose for a full trial usually means switching, sometimes to a different class.
- Add something. Augmentation with a second medication is an option, though the added side-effect burden is a genuine consideration and worth weighing against non-medication options.
- Revisit the diagnosis. Depression that consistently doesn't respond sometimes turns out to be depression plus something else, or something else entirely.
Bipolar spectrum illness, OCD, PTSD, thyroid problems, and sleep apnea all belong on that list.
Write it down, because you'll need it later
Keep a simple record of every medication trial. Not a diary, just the facts:
- Medication name
- Highest dose you reached
- Dates you started and stopped
- Whether you took it consistently
- What happened, in one line
- Why you stopped: no effect, partial effect, side effects, or cost
This is useful the moment you change providers, because the alternative is reconstructing years of treatment history from memory in a twenty-minute appointment.
It becomes far more useful if you ever look at TMS or Spravato. Insurance coverage for those treatments typically depends on documented antidepressant trials that didn't produce an adequate response, and "adequate" is assessed on exactly the dose-and-duration criteria above. People are sometimes told they don't qualify when the real problem is that their treatment history was never written down properly. A list you kept yourself fixes that.
When you've given it long enough
At some point the answer to "how long should I give this one?" becomes "you've given enough of them." If you've completed two or more full trials at proper doses without an adequate response, that meets the working definition of treatment-resistant depression, and it's the point at which interventional treatments become both clinically appropriate and generally coverable.
That's a different conversation from the next prescription, and it's worth having deliberately rather than drifting into year six of trial and error. If you're not sure where you stand, the Find Your Treatment Fit questionnaire is a starting point.
Should I keep taking it if I feel worse in the first week?
Tell your prescriber, don't just endure it. Mild side effects that settle are expected. Worsening mood or new agitation isn't something to wait out.
How do I know if it's working when depression makes everything feel the same?
Rating scales help, which is why prescribers use them. Sleep, appetite, energy, and concentration often shift before mood does, so track those specifically rather than asking yourself the unanswerable "do I feel better.
Can I speed things up by going straight to a higher dose?
That's a prescriber decision. Starting low and titrating up is usually about tolerability, not caution for its own sake, and jumping ahead tends to produce side effects that end the trial early.
If one antidepressant didn't work, will the next one fail too?
Not necessarily, particularly if the next is a different class. But the odds shift with each failed trial, which is why the conversation changes after two.
Does the waiting period restart if I change dose?
Effectively, yes. The meaningful clock is time at the dose you're actually being assessed on.





















