Why psychotherapy hurts — and why that is often the sign it is working

A scene repeats itself with regularity in the first weeks of any serious analysis. The patient begins psychotherapy with a reasonable expectation of relief: he sought help because he was suffering, imagines the process will slowly make the suffering more manageable, and expects to leave each session a little lighter than he came in. Somewhere between the third and eighth session, that expectation is frustrated. He leaves the session worse. He is anxious at night. He has a strange dream. He cries for no apparent reason on Monday morning. He feels irritation toward close people who never used to bother him. And he arrives at the next session with a question he often does not voice: am I getting worse from this? Should I stop?

That is one of the most important questions of the early stages of an analysis, and the clinical answer must be given precisely — because both “yes, this is bad, stop” and “no, hold on, it will pass” can be wrong depending on what is actually happening. This text is about that distinction. And it begins with a thesis that psychoanalysis has sustained for more than a century: an important part of what hurts in analysis is hurting because the work is beginning to function, not because it is failing.

What emerges when the silence loosens

The starting point is understanding what psychotherapy technically mobilizes. Before any analysis, the patient lives with a certain psychic organization that served — well or poorly — to keep life running. That organization includes defenses, avoidances, ways of not feeling what would be costly to feel, ways of not thinking what would be costly to think. What analysis does, when done well, is create a space in which those defenses loosen. Not because the analyst attacks them — a good analyst does not — but because the setting, the regularity, the attentive listening, and the guarantee of confidentiality produce a psychic temperature in which what was contained begins to move.

When that starts to happen, material that had been kept out of consciousness begins to present itself. Not necessarily as organized thought — often as unbound anxiety, as sadness that appears without explanation, as irritation with specific people in the close circle, as dreams that bring old scenes. It is the unconscious delivering material. Freud called this the return of the repressed. André Green described it, among other things, as the work of the negative — what had been maintained in absence begins to present itself.

This material does not arrive comfortable. It arrives whole, with its own weight, and it arrives before the patient has the resources to elaborate it. That is why it hurts. Not because the process is wrong — because the process is beginning to do what it set out to do.

Resistance has a recognizable form

There is a clinical sign that usually accompanies this moment, and it is useful to recognize: resistance to continuing appears precisely when the work begins to access something important. It is as if the psychic structure, perceiving that something it had kept protected is about to be reached, mobilizes to interrupt the movement. Freud described this extensively in texts such as “Remembering, repeating and working through.” The clinical reading he proposed — one that experience with patients keeps confirming — is that the intensity of resistance is often proportional to the importance of the material about to emerge. In other words: when the patient feels a strong urge to quit, it is often because the work is about to reach something decisive.

Not always. And that is where the reading must be done carefully — because not every pain in analysis is the pain of elaboration.

When the pain is a sign that something is wrong

There are situations in which post-session discomfort does not signal the work advancing, but a problem in the setting or in the analytic relationship. It is worth naming these situations, because confusing one with the other has real consequences.

If the patient feels that the analyst intrudes, over-interprets, makes remarks that come across as aggressive or disqualifying, that is not elaboration — it is technical failure. A well-trained analyst knows that premature or ill-timed interpretation produces pain without corresponding clinical yield. If the discomfort has that quality — of intrusion, humiliation, incomprehension — it is worth bringing the matter to the session itself. A good analyst receives that speech without defensiveness and adjusts the technique.

If the patient leaves the session feeling unheard, that the analyst was distracted, that there was no continuity with the previous session, that is not elaboration either — it is a sign that the therapeutic alliance is fragile. A fragile alliance can be worked on, but it cannot be ignored. It is worth talking about.

If post-session pain is accompanied by progressive symptom worsening over months — not transient turbulence, but sustained deterioration — a reassessment is warranted. An analysis that continues to worsen without response to the work for many consecutive months calls for review: to adjust the method, to add pharmacological treatment if indicated, or, in specific cases, to recognize that the analyst-patient combination is not producing the necessary clinical encounter.

What to do when the discomfort appears

The most important clinical recommendation is simple, and it goes against a common instinct: bring it to the session. If you are leaving sessions worse and this is repeating, do not keep it outside the office. Well-trained analysts expect this kind of speech and know it usually opens important work. A sentence like “I left the last session feeling terrible, and I carried it through the week” opens exactly the material the previous session mobilized.

Staying quiet out of politeness, out of fear of offending the analyst, or out of doubt about “whether it’s normal to feel this,” is what usually makes patients quit at moments when a direct conversation would have unlocked weeks of work.

The goal of the process

It is worth ending with what psychotherapy sustains as a real possibility: the goal of the work is not to make the patient suffer. It is that the suffering that already existed — elsewhere in life, in diffuse form — begins to present itself within the setting, where it can be worked on, so it stops commanding the rest.

What hurts in analysis, when it hurts for the right reason, is precisely what had already been hurting underneath, now with a name, with form, with contour. And what gains a name can be elaborated. That is the clinical wager that sustains the method. Not that pain will end — no serious analysis promises that — but that pain ceases to be unnamable and, because of that, ceases to command life in the same way.

If you have been in analysis for a short time and are feeling worse than when you started, it is worth trusting the process a little further — and, above all, talking to the analyst about what has been happening. That is where the answer to whether the pain is signal of work or signal of problem gets built. Not in silent doubt, but in the direct conversation the setting was created to hold.

Dr. Leonardo Sodré is a medical psychiatrist in Brasília (CRM-DF 14,206 · RQE 14,761), with formal training and interest in psychoanalysis. This content is informational and educational, and does not replace individual clinical evaluation, diagnosis, or treatment.