Transference and Countertransference in Therapy and Life
Transference is what happens when feelings from an earlier relationship arrive in a present one: you respond to a therapist, a manager or a doctor partly as though they were someone from your past. Countertransference is the other side of it, the clinician's own reaction to you. This guide explains both, where the ideas came from, how therapy uses them, and how to notice them in ordinary life.
Reviewed byMichael Callans, MSWThe terms, defined
Transference vs countertransference
Whose reaction it is
Transference belongs to the client: the past arriving in how the therapist is perceived. Countertransference belongs to the therapist: their reaction to the client. Both happen in every pairing, and in the relational view they shape each other rather than occurring separately.
What it is used for
Transference gives a live sample of a pattern instead of a report about it, which is why it is treated as material. Countertransference is read as information about what tends to happen around this person: who starts protecting them, bristling at them, or going flat.
What should happen with it
Transference can be named, explored, or left alone depending on whether it helps. Countertransference goes to supervision and the therapist's own therapy first; it is never the client's job to manage, and acting on it is a boundary violation.
What it looks like in real situations
- A manager's two-word reply reads as contempt, and you spend the evening drafting a defense of work nobody criticized.
- You find yourself performing for a therapist met only four times, editing out the week's worst part so as not to disappoint them.
- A calm, competent doctor produces an urge to be the easy patient, and you downplay a symptom to keep that going.
- A new friend's slow reply triggers certainty that you are being dropped, arriving before you check whether they are simply busy.
- Your partner clears their throat the way your mother did, and the argument that follows is pitched at a volume the moment cannot explain.
- A therapist notices they are oddly reluctant to end sessions on time with one particular client, and takes that to supervision rather than extending the hour.
The noticing sheet: five questions after a reaction that felt too big
- What happened, in facts only. What the person actually said or did, in the words they used, with no interpretation attached.
- What I felt, and how strong it was out of ten. Name the feeling rather than the verdict: not 'he was being dismissive' but 'I felt small and angry, about an eight'.
- Who does this person remind me of? Go with the first name that comes, even if the resemblance is only a tone of voice, a posture or a role.
- What did I expect was about to happen? Write the prediction as a sentence: 'I expected to be blamed and then shut out.'
- What actually happened next? Compare it with the prediction. Note whether the expected thing occurred, partly occurred, or did not.
- What does the gap suggest, and what would I do differently if I met the prediction again? If there was no gap, the reaction may be an accurate read rather than a transfer, and that is a legitimate answer.
Common confusions worth clearing up
- Not every strong reaction is transference. Sometimes the other person really did behave badly, and naming your history is a way of avoiding that conversation.
- Transference is not a symptom or a flaw. It is how relational memory works, which is why clinicians treat it as material rather than as pathology.
- A therapist may not use it to deflect a complaint. 'That sounds like your father' is not an answer to 'you forgot what I told you last week'.
- Liking or disliking your therapist is not automatically transference. Fit is real, and so is incompatibility.
- Countertransference is not the client's problem to solve. It is managed in supervision and in the therapist's own therapy.
- Romantic or sexual feeling toward a therapist is common and can be spoken about. Any action on it by the therapist is misconduct and reportable to their licensing board.
What the research actually found
The best-known experiment on this is Høglend and colleagues' 2006 trial in the American Journal of Psychiatry: one hundred patients randomized to a year of dynamic psychotherapy with or without transference interpretations. There was no overall difference between the two arms; both groups improved. The result that stood out was a moderator effect, and it contradicted the team's own prediction. People with a lifelong history of less mature object relations gained more from the transference work, while people with more mature relationships showed small negative effects from it.
Two things follow. First, working explicitly with the relationship in the room is a tool suited to some people and not others, rather than a sign of a deeper therapy. Second, it is a single trial with a counterintuitive moderator finding, and the editorial that ran alongside it asked for replication before anyone builds practice on it. Treat it as the best available evidence and as provisional at the same time.
Related tools & tests
Keep finding the same pattern with new people?
A therapist can work with it as it happens, which is harder to do alone. Free to search our directory.
About this tool
Transference is the arrival of feelings from an earlier relationship inside a present one. A new manager's short email lands as contempt because that is what short messages meant at home. A doctor becomes someone whose approval you need. A therapist who runs five minutes late produces a flare of being unimportant that is out of all proportion to five minutes. You are responding to the person in front of you and also, at the same time, to a template built long before you met them. The word comes from psychoanalysis, where Freud described it in his 1912 paper The Dynamics of Transference, but the phenomenon is not confined to therapy and does not require any particular theory of the mind to be visible.
Freud's own view of it changed, and that history explains some of the confusion around the term. At first he treated transference as an obstruction: an intrusion of the past that got in the way of the work. By 1912 he had decided it was the work, the place where a pattern could be observed live rather than merely reported. His framing was displacement of feelings from childhood figures onto the analyst. Most contemporary clinicians describe it more broadly: everyone carries internal expectations about what relationships do, formed in the relationships that taught them, and those expectations shape perception of every new person. That account connects transference to attachment patterns and to what cognitive therapists call schemas.
The relational turn in psychoanalysis added something important. If transference were purely the patient's distortion, the therapist could sit outside it as a neutral screen. The relational view, associated with writers such as Stephen Mitchell and Jessica Benjamin, holds that it is co-created: a therapist who is in fact slightly withholding will activate a withholding-parent template more readily, and the pattern that appears in the room belongs to both people. This matters practically, because it rules out the move where a clinician explains away a reasonable complaint as transference. A reaction can be shaped by your history and be an accurate read of what the other person is doing.
Read the diagram as text
Two people face each other in solid ink. Behind each of them stands a dashed figure drawn with the same outline, offset up and outward: on your side the therapist's own earlier figure, on theirs someone from your past. A brand path arcs over the gap from you and lands on the dashed figure behind the therapist, not on the therapist. A second brand path arcs back low from the therapist, passes under both of them and rises into the dashed figure behind you. The brand marks the feeling in transit and where it settles. On the phone canvas the two pairs stack, with one brand path running down the right and the other back up the left.
Where the feeling actually lands
Conceptual route, not counts
The earlier relationship: What a short message, a late arrival or withheld approval meant back then
The template it left: An expectation about what relationships do, carried into every new one
The person in front of you: A therapist, a manager, a doctor, a partner, met for the first time
The reaction: Out of proportion to what happened, arriving fully formed, oddly familiar
The clinician's own side: Their earlier figures are in the room too, so the pattern belongs to both people
The earlier relationship → leaves behind → The template it left
The template it left → is laid over → The person in front of you
The template it left → the feeling answers the template, not the person → The reaction (error transition)
The person in front of you → pulls a response from → The clinician's own side
In therapy, transference shows up in recognizable shapes. Positive transference: the therapist becomes unusually wise, kind, the parent you did not get, and you work hard to be a good client. Negative transference: you expect criticism, hear it where it was not, go quiet, come late, consider stopping. Erotic transference: romantic or sexual feeling toward the therapist, which is common, usually bewildering, and not a sign that anything has gone wrong. Many people also experience a mix that shifts week to week. The setting intensifies all of it on purpose: regular times, close attention, asymmetric disclosure, and a person who does not retaliate.
Outside therapy, the same thing runs the background of ordinary life. Bosses get fathers; mentors get the parent who paid attention; the partner who sighs gets the parent who sighed. The useful tells are not mysterious. The intensity is out of proportion to what actually happened. Your response arrives fully formed, before you gather much information. There is a strong sense of familiarity about a person you barely know. And the script is one that has run before, with the same ending, across several unrelated people. One of those signs alone means little. All four together usually mean the present person is standing in for someone else.
Countertransference is the clinician's side. Freud introduced the term around 1910 and treated it mainly as a contamination to be analyzed away in the therapist's own treatment. Paula Heimann's short 1950 paper On Counter-Transference changed the field's mind by arguing that the analyst's emotional response to the patient is one of the most important tools of the work: if you feel uncharacteristically bored, protective, irritated or helpless with a particular person, that feeling is data about what happens around that person. Glen Gabbard's contemporary model, set out in 2001, treats it as jointly created: partly the therapist's own history, partly a response pulled by this patient's way of relating, and separating the two strands is a core clinical skill.
What a therapist does with it is mostly unglamorous. They notice the pull, name it to themselves, and take it to supervision or their own therapy rather than acting on it. Sometimes they use it directly: a therapist who notices they have started softening every piece of feedback might wonder aloud whether other people in your life also handle you carefully, and what that costs you. Sometimes they say nothing and simply decline to play the part being offered, which is often the more useful move. What they should never do is make their feelings your responsibility, and acting on romantic or sexual countertransference is a serious boundary violation reportable to a licensing board, not a gray area.
The evidence on how much to work with transference in the room is more interesting than the usual confident claims. Høglend and colleagues ran a randomized trial, published in 2006, in which one hundred patients received a year of dynamic psychotherapy either with or without transference interpretations. Overall, the two arms came out the same: both improved, with no main effect for the transference work. The finding that mattered was a moderator, and it ran against the researchers' own hypothesis. Patients with a lifelong pattern of less mature object relations did better with transference interpretations, while those with more mature relationships showed small negative effects. The accompanying editorial called for replication, which is the right caution to carry.
The practical reading of that trial is that transference work is a tool matched to a person, not a mark of depth. A therapist who routes everything back to the relationship between the two of you is not thereby doing better therapy, and one who rarely mentions it is not shallow. If you find that the focus on the relationship in the room leaves you feeling examined rather than helped, that is worth saying, and a competent therapist will treat your reaction as information rather than as resistance to be interpreted. The same goes in the other direction: if you keep having a reaction to your therapist that you never mention, the mention is usually where the useful part starts.
Two final clarifications. Not everything is transference. Sometimes a therapist really is dismissive, really did forget what you told them, really is not a good fit, and dressing that up as your history is a way of avoiding an ordinary conversation about it. And transference is not a symptom or a defect; it is how memory works in company, which is why it is better treated as material than as a problem. The noticing sheet on the printable below is built for one reaction at a time: who the person reminded you of, what you expected, what they actually did, and what the gap suggests. It pairs well with What Are Schemas?, which maps the templates themselves, and Inner Child Work, which goes at the early relationships that formed them.
References
- Freud S. The dynamics of transference (1912). In: The Standard Edition of the Complete Psychological Works of Sigmund Freud, Volume 12. London: Hogarth Press; 1958.
- Heimann P. On counter-transference. International Journal of Psycho-Analysis. 1950;31:81-84.
- Gabbard GO. A contemporary psychoanalytic model of countertransference. Journal of Clinical Psychology. 2001;57(8):983-991.
- Høglend P, Amlo S, Marble A, et al. Analysis of the patient-therapist relationship in dynamic psychotherapy: an experimental study of transference interpretations. American Journal of Psychiatry. 2006;163(10):1739-1746. doi:10.1176/ajp.2006.163.10.1739
- Gabbard GO. When is transference work useful in dynamic psychotherapy? American Journal of Psychiatry. 2006;163(10):1667-1669.
- Hayes JA, Gelso CJ, Hummel AM. Managing countertransference. Psychotherapy. 2011;48(1):88-97.
Cite this source
Psychology.com. (2026, November 6). Transference and Countertransference. Psychology.com. https://psychology.com/tools/transference-and-countertransference
Transference and Countertransference FAQ
What is transference in simple terms?
Transference is responding to someone in the present as though they were someone from your past. The feelings, expectations and reflexes built in an earlier relationship show up in a current one, so a manager's blunt email can produce the dread a parent's blunt tone once did, or a therapist's warmth can be met with the wariness that warmth used to require. Freud described it in 1912 in the consulting room, where it is strongest, but it happens with bosses, doctors, teachers, landlords and partners. The clearest sign is a reaction that arrives fully formed and is out of proportion to what actually occurred.
What is countertransference?
Countertransference is the therapist's emotional reaction to the client. Freud first treated it as interference to be removed, but Paula Heimann's 1950 paper argued the opposite: that the therapist's response is one of the most valuable instruments they have, because feeling unusually protective, bored, irritated or helpless with a particular person is information about what happens around that person. Glen Gabbard's contemporary model treats it as jointly produced, part the therapist's own history and part a reaction this client tends to pull from people. It belongs in supervision and the therapist's own therapy, never on the client to manage.
What is the difference between transference and countertransference?
Transference is the client's: the past arriving in how they see and respond to the therapist. Countertransference is the therapist's: their reaction to the client. The direction is the whole distinction, though in the relational view the two shape each other rather than running on separate tracks, and when they interlock and the old pattern actually plays out between them, clinicians call that an enactment. The handling differs too. Transference may be named and explored in the room; countertransference is worked on outside it, and acting on it, especially romantically or sexually, is professional misconduct.
Is transference a bad thing?
No. Transference is treated as material rather than as a malfunction, because it brings a relational pattern into the room live instead of leaving it as a story you report about other people. That is exactly what makes it useful: the thing can be examined while it is happening, with someone who will not play the expected part. Freud moved from seeing it as an obstacle to seeing it as the central lever of the work, and that remains the mainstream position. What is unhelpful is a therapist using it to explain away a legitimate complaint, or a client using it to decide something about themselves is broken.
I think I am developing romantic feelings for my therapist. What should I do?
This is common, it has a name, erotic transference, and it does not mean the therapy has gone wrong or that you did anything wrong. The setting is built for it: regular attention, no retaliation, and a person who knows a lot about you while revealing little. The most useful thing is usually to say it, awkward as that is, because a trained therapist will treat it as material and the conversation itself often reveals what the feeling is reaching for. What should never happen is the therapist acting on it or encouraging it. That is a boundary violation, reportable to their licensing board, and ending the work is a reasonable response.
How do therapists handle their own reactions to clients?
Mostly outside the session. They notice the pull, bring it to clinical supervision or consultation, and work on the parts that are clearly their own in personal therapy, which is one reason most training programs require both. Inside the session, the use is indirect: a therapist who notices they have begun handling someone with unusual care might ask whether other people do the same and what that costs, without making their own feeling the topic. Declining to play the offered role is often the more useful move. What is not acceptable is venting at a client or asking a client to look after them.
Does working with transference make therapy more effective?
Not reliably, and the best test of this is more nuanced than either side's slogans. Høglend and colleagues randomized one hundred patients in 2006 to a year of dynamic psychotherapy with or without transference interpretations and found no overall difference: both arms improved. The notable result was a moderator that ran against the team's own prediction, with patients whose lifelong relationships were less mature benefiting more from transference work and those with more mature relationships showing small negative effects. The reasonable conclusion is that it is a tool matched to a person rather than a measure of therapeutic depth, and that the finding still needs replication.