In the first week of TMS treatment, almost every patient I see asks some version of the same question: when am I supposed to feel different?

The version I hear most often comes from people who have spent months or years cycling through antidepressants, where "when will I feel different" became a kind of background ache calibrated to the slow eight-to-twelve-week timeline of SSRI trials. The TMS course is shorter than that — but the timing inside it is more specific, and less linear, than the way most patients picture it before they start.

This article is the realistic treatment timeline I give patients during their first consultation — what happens when, what tends to change first, when the strongest improvement typically arrives, when to know the treatment is working, and when to know it isn't. The headline answer is "between weeks two and four for most patients" — but the more useful answer is the texture inside those weeks.

What "working" means in TMS — a quick framing

Before walking through the timeline, two terms are worth being precise about, because they answer two different questions.

Response means a 50% or greater reduction in depression symptoms by the end of the treatment course. It's the standard clinical threshold for "the treatment is helping."

Remission means the symptoms have come down to a level that no longer meets criteria for depression — the patient is, in clinical terms, well.

Across published trials, response rates in treatment-resistant depression sit around 50–60%, and remission rates around 30–40%. (I covered the response and remission rates in more detail in a separate article.) The timeline below is the path most responders walk through to get there.A meaningful minority of patients — around 30–40% — don't reach response. The timeline still matters for them, because it determines when we know to change course rather than keep going.

The standard course structure

 A conventional rTMS course for depression is five sessions per week for four to six weeks, totalling 20 to 36 sessions. Each session is around 19–25 minutes for standard rTMS, or about 3 minutes for newer theta-burst protocols.

Most of the timing patterns below apply to this structure. Accelerated and intensive protocols compress the schedule — more on those further down — but the underlying biology of network reorganization doesn't compress proportionally, and the timeline of noticeable change still follows a recognisable arc.

The realistic week-by-week timeline

Sessions 1–5 (Week 1) 

Most patients feel essentially nothing mood-wise in the first week, and this is normal. The first week is when side effects in the first week — mild scalp tenderness, occasional headaches, transient fatigue from the new routine — are most noticeable. By the end of the week, most of these have begun to settle.Patients sometimes report a faint sense of "something is different" by session four or five — slightly easier mornings, a brief patch of energy that wasn't there before — but these are usually small enough to dismiss. They aren't the response itself. They are early signals that the network is registering the stimulation.

What I tell patients in week one: the absence of mood change right now is not a sign the treatment isn't working. Almost no one experiences clinical change this early.

Sessions 6–10 (Week 2)

This is the window where the first reliable early signals tend to appear. The order matters, because it's almost always the same.

Sleep usually shifts first. Patients describe falling asleep more easily, waking less often through the night, or — if they had been oversleeping — needing less sleep and waking more rested. Sleep change is the earliest indicator that the treatment is engaging the system it's supposed to engage.

Energy is the next to shift. Patients describe being able to get through a day without the late-morning collapse, or making it to evening without needing a nap. Motivation often hasn't returned yet, but the bandwidth has.Mood typically follows. This is the part patients are waiting for — and it's also usually the third thing to change, not the first. Patients who fixate on mood as the only marker of progress sometimes miss the sleep and energy shifts that precede it, which is part of why we ask about all three at every weekly review.

Sessions 8–12 (Week 2–early Week 3)

This is the common window for the TMS dip — a temporary worsening of mood, energy, or sleep that shows up in a meaningful subset of patients (around 60–70% in my practice) and lasts roughly 5–10 days. The dip is not failure. What comes after the dip is usually the most pronounced improvement of the course. I've written about it in detail in a separate article — the short version is: flag it to your clinician, keep showing up, expect the dip to resolve.

The exception, worth flagging in every article on this topic, is new or intensifying suicidal thoughts. That is not the dip and requires same-day contact with your treating clinician, not waiting it out.

Sessions 13–18 (Week 3–early Week 4)

For most responders, this is the window where improvement becomes hard to ignore. Patients describe sleeping consistently better, having stretches of normal energy through the day, and noticing that small things they used to dread — a phone call, leaving the house, opening a backlog of emails — feel less heavy.

Mood usually arrives in a quieter way than people anticipate. Patients describe it as "I noticed I laughed at something on the radio," or "I made plans this weekend without forcing myself," or "I caught myself looking forward to something." The dramatic-recovery scene that some patients are bracing for almost never materialises — and that's a good outcome, not a disappointing one. The gradual version is the one that holds.Family members often notice before the patient does. Patients sometimes get hard data this way — a spouse or sibling saying, "You've been different the last few days, did you notice?" — and only then recognise the shift themselves.

Sessions 19–30 (Week 4–6)

This is the consolidation phase. Improvement that began in week three continues to deepen. For patients who are responding, week five or six is typically when the depression score crosses from "much better" to "well." For patients who are partial responders, this is when we have the most useful clinical data to decide whether to extend the course, switch protocols, or augment with another treatment.What I tell patients in this window: if you're feeling significantly better, this is the part of treatment to protect — keep coming in, complete the course as planned, don't taper early. The improvement is most durable when the full course is finished.

Why the order of recovery is sleep → energy → mood

This is one of the most consistent patterns in TMS, and it isn't accidental. Sleep architecture is regulated by circuits that overlap with the depression network but respond to changes in connectivity faster than the broader mood-regulation system. Energy follows because energy is, in part, downstream of restored sleep. Mood is the slowest because the cortical and limbic systems that produce subjective mood states require the most network reorganization before the felt experience shifts.

This ordering is useful for two reasons. First, it gives patients a way to track progress in week two before mood has moved at all. Second, when one of these markers shifts and the next doesn't follow, it's an early signal that something in the protocol may need adjusting.

When you know it isn't working

The honest answer is that the picture usually becomes clear by the end of week four. A patient who reaches session 20 with no shift in sleep, no shift in energy, no quiet positive signals from anyone around them, and no fluctuation that could even be a dip is, statistically, unlikely to convert into a strong responder in the final two weeks.

That doesn't mean treatment is over. It means we have several options to discuss at the four-week review:

Protocol adjustment — changing stimulation intensity, target location, or pulse frequency

Switching protocols — for example, moving from standard left-sided rTMS to bilateral, or to a different evidence-supported target

Adding theta-burst or another protocol variant

Augmenting with another modality — therapy intensification, ketamine, or a medication change

Moving to a different treatment entirely if neuromodulation isn't the right fit

Non-response on TMS is not a clinical dead end. It is information that helps direct the next step.

Protocol differences — accelerated and intensive courses

Several newer protocols compress this:

Theta-burst stimulation (iTBS) delivers the same total dose of stimulation in roughly 3 minutes per session instead of 19 minutes. The course length is similar — typically four to six weeks — but each individual session is much shorter. Response timelines closely track conventional rTMS.

Accelerated TMS delivers multiple sessions per day, often across one to two weeks, to compress the full course into a shorter calendar window. Some protocols deliver up to ten sessions per day. The published timeline of response in these protocols is faster in calendar terms — patients can reach response within one to two weeks — but the underlying mechanism is the same network reorganization, just driven harder in less time. These protocols are useful in specific clinical situations, particularly when a slower course is impractical, but they require a clinic equipped to deliver them safely.

Deep TMS (dTMS) uses a different coil geometry to reach deeper cortical regions, and is FDA-approved for major depression and OCD. Timeline of response is broadly comparable to standard rTMS, though some patients with certain symptom profiles may respond differently.

The right protocol is a clinical decision made on the basis of diagnosis, prior treatment history, and practical considerations — not a marketing differentiator.

What happens after the course ends

For patients who reach response or remission, the next question is durability. The published data suggests that the average TMS response holds for several months in many patients, with some experiencing benefit lasting a year or longer without further treatment. A subset of patients require maintenance sessions — typically a small number of additional sessions every one to three months — to sustain the gains.

A relapse-prevention plan is part of finishing a TMS course well. This usually includes ongoing therapy, lifestyle structure (sleep, exercise, daylight), continued medication where appropriate, and a clear plan for what to do if early relapse signs appear. Patients who actively maintain these supports tend to hold their gains longer than patients who treat TMS as a one-time fix.

Common patient frustrations during the timeline

A few of the most frequent reactions during a TMS course, and what's actually going on:

"It's been a week and I don't feel any different." Normal. Mood change in week one is rare. Sleep and energy are the markers worth tracking in this window.

"I felt good in week one and now I feel worse." Often the TMS dip. Time-limited, expected, usually followed by the strongest improvement of the course.

"My family says I'm better but I don't feel it yet." Also common. Mood recognition often lags external recognition by a few days to a week. Trust the people around you in this window — they're not flattering you, they're seeing something real."

I missed a session. Did I ruin it?" No. Missing one or two sessions occasionally does not derail a course. Missing many consecutive sessions, or stopping mid-course, is what causes the most preventable failures.

The bottom line

For most patients on a standard rTMS course, the realistic timeline looks like this:

Week 1: no mood change yet. Side effects appear and begin settling.

Week 2: early signals — sleep first, then energy.

Week 2 or 3: a possible temporary dip in a meaningful subset of patients.

Week 3 to 4: improvement becomes hard to ignore for responders.

Week 4 to 6: consolidation. The four-week mark is the most useful clinical decision point.After the course: durability for many patients, maintenance sessions for some.

The single most useful framing I can offer is this: the question is not when TMS will work, but whether you and your clinician are watching the right markers in the right order. Sleep first, energy next, mood last. Most patients who fixate on mood from session one mistake the early signals for nothing — and then are surprised in week three when they realise the change has been building for two weeks already.If you are weighing TMS as a treatment option, the most useful next step is a structured clinical evaluation with a psychiatrist who works with neuromodulation. That conversation should cover whether TMS is the right protocol for your diagnosis, which course structure makes sense, what realistic outcomes look like for your history, and how progress will be measured.

Common questions, briefly answered

What's the fastest TMS can work? Some patients notice the earliest signals (sleep, energy) by sessions 4–6, but reliable mood change before week two is uncommon on standard protocols. Accelerated protocols can compress the timeline.

Can I know after one or two sessions whether it will work? No. The biology of network reorganization takes weeks. Early sessions establish the protocol and the threshold, not the outcome.

How will I know if it isn't working? By the end of week four, the clinical picture is usually clear enough to make a decision about extending, adjusting, or switching. A flat line across all three markers (sleep, energy, mood) by session 20 is the signal to review.

How long does the improvement last after the course? Variable. Many patients hold the response for several months to a year. Some need maintenance sessions every one to three months. The strongest predictor of durability is finishing the full course and keeping the basics (sleep, therapy, lifestyle) in place afterward.

Is faster always better? No. Accelerated protocols are useful in specific situations but are not inherently better. The right schedule is the one that matches your diagnosis, your life, and your clinician's clinical judgement.

Does TMS for OCD have the same timeline? Roughly similar, with some differences in the typical response window. I'll cover the OCD-specific picture in a separate article in this series.