Countertransference In Therapy

Countertransference is a psychological phenomenon that occurs when a clinician lets their own feelings shape the way they interact with or react to their client in therapy.

Often, both the clinician and the client are unaware it is happening.

A client can unconsciously remind a therapist of someone from their own life, prompting a biased reaction the therapist must stay alert to (Overstreet, 2021).

What Does Countertransference Look Like?

Countertransference occurs whenever a therapist brings in their own experiences to the extent that they lose perspective of the client’s own and stop being objective (Jacobson, 2022).

Examples of countertransference include when the therapist:

• Over-identifies with the client’s stories and shares too many about themselves
• Offers a lot of advice instead of listening to the client’s experience
• Pushed the client to take action the client doesn’t feel ready for
• Wants to relate outside of the therapy room
• Inappropriately disclosed personal information
• Develops romantic feelings for the client
• Does not have adequate boundaries with the client
• Is overly critical or supportive of the client

Transference vs. Countertransference

Transference and countertransference describe the same therapeutic relationship from opposite directions.

• Transference: the client unconsciously redirects feelings from a past or present relationship, such as with a parent, partner, or friend, onto the therapist.
• Countertransference: the therapist’s own emotional reaction to the client, including to that transference.

Both processes are usually unconscious. Transference is an accepted, even useful, part of therapy: a skilled therapist can read it to understand a client’s key relationships. Countertransference, by contrast, is the clinician’s own response to that transference (Overstreet, 2021).

It is normal, but clinicians are responsible for monitoring it. Left unchecked, it can compromise their objectivity and harm the client.

Warning Signs of Countertransference

How can a therapist or client tell if countertransference is affecting a session?

Warning Signs in Adult Therapy

Both the therapist and the client should watch for warning signs. An inappropriate emotional response from the therapist is one: an unreasonable dislike for the client, excessive positive feelings, over-emotionality, or thinking about the client between sessions.

Dreading the session, or feeling uncomfortable during it, is another red flag for either person.

Warning Signs in Child Therapy

Countertransference can also appear in child therapy. Warning signs on the therapist’s side include ignoring a child’s harmful or unusual behavior. Fantasizing about rescuing the child from their situation is another. So is encouraging the child to act out (Fritscher, 2021).

Impact of Countertransference on Therapy

As described, countertransference can appear in many ways. Most presentations, though, are harmful to the relationship and can have adverse effects.

For instance, a clinician who brings personal experience or emotion into a session can lose perspective and end up hurting the client. Erotic countertransference is one especially concerning form.

This is when the clinician develops attraction, love, or sexual feelings toward a client, which is prohibited and carries strict ethical and legal consequences (Overstreet, 2021).

Not every effect is negative, though. Some forms of countertransference are genuinely beneficial.

Meta-analytic evidence supports this: successfully managing countertransference, not simply gaining more experience, predicts better therapy outcomes (Hayes et al., 2018; see Critical Evaluation below).

A Brief History of Countertransference

Freud’s Classical View

Sigmund Freud introduced the term “countertransference” in 1910. He wrote that the analyst must “recognize this counter-transference in himself and overcome it” (Freud, 1910).

Freud treated it as the analyst’s own failing, not as useful information to act on. He tied this directly to the analyst’s own training analysis, requiring the analyst to have resolved their own conflicts before treating patients (Freud, 1910).

From Failing to Clinical Tool

This began to change in the 1950s, inside the object-relations tradition associated with Melanie Klein. Paula Heimann argued that the analyst’s whole emotional response to the patient is “one of the most important tools for his work,” not a problem to eliminate (Heimann, 1950).

Donald Winnicott took this further.

In “Hate in the Counter-Transference,” he argued that a therapist can genuinely come to hate a very difficult patient. He treated that reaction as normal, not a personal failing (Winnicott, 1949).

Heinrich Racker gave this idea a clinical structure.

In concordant identification, the therapist feels something like the client’s own feelings. In complementary identification, the therapist instead feels like someone else from the client’s life, such as a critical parent (Racker, 1957).

Joseph Sandler pushed this a step further.

He argued that a client unconsciously pressures the therapist, through subtle cues, into taking up a role that fits the client’s own relationship patterns (Sandler, 1976).

Not every analyst welcomed the change.

Otto Kernberg named a split between two views. The classical view treats countertransference as the analyst’s own unresolved conflicts intruding on the work; the newer view treats it as the analyst’s entire emotional response (Kernberg, 1965).

Countertransference is therefore no longer only a fault to fix. It is also treated as a clinical clue, as long as the therapist can tell their own history apart from what the client is actually evoking.

Critical Evaluation of Countertransference Research

Since the 1950s, research has moved counter-transference from an idea supported mainly by clinical case material toward one with a genuine evidence base, though real limits remain.

Contemporary Research

The clearest modern test of whether managing countertransference actually helps therapy comes from a set of meta-analyses by Hayes and colleagues.

• Aim: To combine the existing research on the links between countertransference, countertransference management, and therapy outcome.
• Method: Three random-effects meta-analyses pooled published and unpublished studies on countertransference reactions, management factors, and outcome. Sample sizes ranged from 9 to 14 studies across the three analyses (Hayes et al., 2018).
• Results: Raw countertransference reactions were modestly linked to worse outcomes (r = -.16). Management factors significantly reduced countertransference reactions (r = -.27). Successful management was linked to better outcomes even more strongly (r = .39) than either of the other two effects.
• Conclusion: Left unmanaged, countertransference is a modest liability, but managing it well is one of the more robust predictors of good therapy outcome found so far (Hayes et al., 2018).

The clearest evidence for this comes from Betan and colleagues (2005), who built the field’s first systematic measure of countertransference.

• Aim: To develop and validate a measure of clinicians’ countertransference reactions, and test whether it relates to a patient’s personality pathology.
• Method: A national sample of 181 psychiatrists and psychologists completed personality-pathology measures and a new Countertransference Questionnaire about one current patient (Betan et al., 2005).
• Results: Factor analysis produced eight coherent countertransference patterns, independent of the clinician’s own theoretical orientation, each linked in predictable ways to the patient’s personality pathology.
• Conclusion: Countertransference reactions can be measured in psychometrically sound ways and are systematically patterned by a patient’s psychopathology across therapeutic approaches (Betan et al., 2005).

More recent research sharpens this picture and tests it well beyond the psychodynamic consulting room.

Does Management Really Predict Outcome?

Tishby and Wiseman (2020) followed 27 therapists treating 67 clients in year-long psychodynamic therapy. They coded each therapist’s characteristic countertransference pattern from recurring parental narratives.

Negative patterns hurt the alliance: more ruptures, less repair. Positive patterns only helped when they reflected a genuinely positive relationship, not an attempt to fix a negative one through the client (Tishby & Wiseman, 2020).

Where the Concept Applies Now

Michaud and colleagues (2021) reviewed the evidence on suicidal patients.

Clinicians often feel disinterest, anxiety, or rejection toward them, and this reaction predicts the patient’s own later suicidal thoughts and behaviour. The review treats countertransference monitoring as a patient-safety issue, not just a technique refinement (Michaud et al., 2021).

A Jointly Created Process

Gabbard (2020) argues countertransference is created jointly, not felt by the therapist alone. A patient can unconsciously pressure the therapist, through projective identification, into acting out a role from the patient’s own inner world (Gabbard, 2020).

This evidence base has two real limits, though.

1. Correlational, not causal: the countertransference-management link has never been tested with a randomized trial.
2. Self-report only: the main measurement tool asks clinicians to rate their own private reactions.

Correlational, Not Yet Causal

Hayes and colleagues themselves flagged significant differences across the studies pooled in all three of their meta-analyses (Hayes et al., 2018), a common issue in psychotherapy-process research generally. None of the effects they report are experimental.

The management-outcome finding alone rests on just nine pooled studies and fewer than 400 participants, a small base for such a broad claim. This matters because therapists who manage countertransference well may simply be more skilled clinicians overall.

The claim that managing countertransference specifically improves outcome, rather than just marking general competence, remains unproven. Correlational designs cannot separate the two explanations.

The clearest next step would be a controlled trial. Researchers could randomly assign trainee therapists to structured countertransference-management training or to standard supervision, then compare client outcomes between the two groups (Betan et al., 2005; Hayes et al., 2018). That would show cause, not just correlation.

A Self-Report Measurement Problem

The field’s main measurement tool has a built-in blind spot. The Countertransference Questionnaire asks clinicians to rate their own cognitive, emotional, and behavioral reactions to a client (Betan et al., 2005).

No outside observer can independently confirm that a therapist’s reported reaction genuinely reflects the client’s inner world rather than the therapist’s own history. This blind spot is not new.

Freud built the whole concept around a phenomenon that, by definition, only the analyst can observe, decades before anyone tried to measure it systematically.

Different studies in the evidence base may not even be measuring quite the same private experience, which makes the concept hard to pin down. Every self-report tool in this field shares the same limitation, psychodynamic or otherwise.