Almost inevitably, at some point in our work as therapists, we’ll have a strong response to a client that seems confusing, uncomfortable or disproportionate. We might find our thoughts circling back to them after the session, notice an uncharacteristic irritation, or feel protective towards them in a way that surprises us. Perhaps we dread a particular session each week without being able to pinpoint why, feel strangely inadequate during and afterwards, or experience a real sense of loss when a particular client cancels. Sometimes the response is less obvious: when we work much harder than usual for a particular client, prepare differently for their sessions, or become reluctant to end on time. We can experience these responses as emotions, thoughts and bodily or somatic sensations, or changes in what we find ourselves doing.
If our core training was process-oriented, we may already have a language for this. A strong response may then seem less alarming: something to notice, become curious about and bring to supervision. If our training was more instrumental than relational, strong on technique or theory but less explicit about what to do with our own internal reactions to a client, we may not be prepared for these experiences at all. Even within programmes that include dedicated teaching on transference, countertransference and intersubjectivity, the ways these concepts apply in practice can be difficult to imagine until we encounter them ourselves. And even experienced therapists can find themselves wondering whether something’s gone badly wrong, whether they’ve somehow breached a boundary, or whether they’re no longer cut out to be doing this work. Meanwhile, most coach trainings include nothing to prepare coaches for this sort of experience either. This post is particularly relevant for qualified and trainee therapists and coaches who are trying to make sense of strong responses to particular clients.
A strong reaction to a client doesn’t necessarily mean that something’s gone wrong. But it does demand our attention. The question is, what might we do with that attention, and what might our reaction tell us?
Starting with curiosity
My starting point, both in my client work and as a supervisor, is curiosity. What information might this difficult experience hold about our own process, the therapeutic relationship, the client’s experience, and/or the wider field conditions? These possibilities are not mutually exclusive, and it’s important that we resist the temptation to make them so even if doing so feels easier.
When I experience a strong response to a client, I might not enjoy the discomfort itself, but I genuinely enjoy grappling with it, both on my own and with my supervisor. What’s the texture of this particular feeling? Is it familiar to me from my personal or professional life? Is something in my own history being activated by this client or by the material they’re bringing? Might I be picking up something that the client is communicating non-verbally? If I am feeling irritated, for example, might the client also be experiencing irritation, whether towards me, towards therapy, or towards someone or something else?
And then there’s the question of what might be happening between us. Is something being co-created in the relationship that deserves to be named rather than continuing to run anonymously underneath the work?
I find questions like these much more fruitful than trying to arrive quickly at a single explanation. They allow us to take the response seriously without assuming that we already know what it means.
Plenty of theories
There are, of course, plenty of theoretical frameworks that help us understand therapists’ strong emotional responses to clients. When I started writing this article, I initially produced a long section acknowledging them and stopped at eighteen. That’s the bones of a book, not a blog post, and not necessarily one you’d want to read, so I’ll stay with what matters most here.
The concept most likely to come to mind in this regard is countertransference. Put simply, ‘transference’ describes something fairly ordinary: we bring what we’ve learnt about relationships into new ones, including therapy, and a client’s response to us will sometimes owe as much to someone or something from their past as it does to what’s happening between us in the here and now. ‘Countertransference’ moves in the other direction: the therapist’s response to the client.
The history of the concept is useful because it helps explain why therapists can become preoccupied with locating the source of a strong response: is it ‘mine’ or ‘theirs’? For the first several decades of psychoanalysis, countertransference was understood largely as the analyst’s unresolved material intruding on the work (Freud, 1910). Later theorists began to consider whether the therapist’s response might also provide information about the client’s unconscious processes (Heimann, 1950). Donald Winnicott (1949) distinguished between personal reactions rooted in the therapist’s own history and what he called ‘objective countertransference’: an understandable, even necessary, response to who the client actually is and how they behave. Heinrich Racker (1957) took this further. In a ‘concordant identification’, which underpins empathy, you recognise something of your own experience in theirs. In a ‘complementary identification’, you find yourself responding as if you were someone from the client’s relational history, such as a critical or emotionally unavailable parent. Later relational approaches placed increasing emphasis on what is happening relationally between therapist and client (Clarkson, 1990; Gelso and Hayes, 2007).
The picture is therefore complex: a strong response may tell us something about ourselves, the client, the therapeutic relationship, or some combination of the three.
In practice, a strong response to a client is rarely an either/or matter. Deciding too quickly that a feeling belongs entirely to the client can become a way of not looking at ourselves, allowing us to set the feeling aside in relief rather than exploring it. Conversely, deciding too quickly that a feeling is entirely our own can shut down curiosity just as effectively, treating the reaction as something to bracket as irrelevant to the therapy rather than potential information about the client and what’s happening between us. Either way, we can end up with a convenient tidiness instead of understanding.
This is why I prefer to hold ‘what might belong to me?’ as an open question rather than something that requires an immediate answer. It can be useful to explore what the reaction touches in us without assuming that it explains the whole experience.
Erskine and the difference between reacting and responding
Richard Erskine’s work is particularly helpful in this respect, partly because he writes so candidly about his own experience of countertransference (Erskine, 2023). He describes carrying considerable shame in his early years as a therapist when clients stirred strong feelings in him. Believing that having a significant emotional response was evidence that something was wrong with him as a practitioner, he kept some of these reactions to himself rather than bringing them to supervision.
Some of his case examples are striking precisely because the eventual meaning he came to was so far from what he initially feared. In one, the unaccountable ‘repulsion’ he felt towards a client finally made sense as an accurate resonance with an experience for which she hadn’t yet found words: childhood memories of her mother’s disgust towards her. In another case, a somatic response that he found deeply troubling acquired a different meaning when the client’s history of childhood sexual abuse emerged.
These vignettes don’t tell us that a therapist’s response is always accurate. In fact, they demonstrate the opposite: it can take time, context and careful reflection to work out what a response is telling us, and the therapist’s first instinct or interpretation may be wrong.
I especially like Erskine’s shift of the question away from where the response originated and towards what it’s actually doing in his framing of reactive and responsive countertransference.
A responsive countertransference remains in the service of the client, whatever its origin. Our own history may shape a moment of tenderness, protectiveness or concern without impeding our ability to be present and to attend to what the relationship requires. A reactive countertransference, by contrast, is when our own needs, conflicts or unfinished business become dominant. This means that a reactive countertransference has stopped serving the client and become about the therapist instead. It’s something that’s solely ours to address, not because we’re at fault for having it, but because we’re the only person in the room who can do anything with it.
This is an important distinction because it means that a response partly rooted in our own history is not automatically something that needs to be eliminated. What matters is whether we can use our experience while remaining oriented towards the person in front of us.
Why this so often goes unsaid
If you’ve felt ashamed of a strong reaction to a client and said nothing, you’re in good company, including Erskine himself. Recent research on why therapists withhold material from supervision consistently found that negative feelings towards clients and countertransference reactions are among the things practitioners are most likely to keep to themselves, not simply because they fear judgement, but because they’re making a real-time assessment of whether a particular supervisor feels safe enough to hear it (Apostol et al., 2025). That’s a reasonable response to a genuinely vulnerable disclosure, and it points to something supervisors, myself included, are responsible for creating the conditions for, not something supervisees should simply push through.
A strong response can contain more than one truth
Imagine a client who regularly questions whether the therapy is helping. They ask whether you really understand them, whether therapy is making any difference, and whether they might be better off seeing someone else. You notice that you feel increasingly inadequate. You begin preparing more carefully for their sessions, thinking about them between appointments, reading more and offering more suggestions in your attempt to demonstrate that you understand what they need.
One might conclude that the client is expecting too much, or that your own fear of failure has been activated. Both may be true.
You may have a particular sensitivity to feeling that you’ve disappointed someone. That vulnerability might make the client’s questioning harder for you to tolerate than it would otherwise be. At the same time, the client’s questioning may be part of a genuine relational pattern. Perhaps their difficulty in trusting that another person will remain interested in them leads them repeatedly to test that interest. Maybe dissatisfaction is how they communicate a need that they find hard to express more directly. Or perhaps they question the therapy because something about it genuinely isn’t working for them.
The fact that we feel inadequate doesn’t in itself tell us which explanation is correct.
I might recognise something personal in my response and still remain curious about what the client is communicating. I can acknowledge that the client genuinely evokes something in me and still take responsibility for what I do with that experience. The personal and the relational can be intertwined without either one cancelling the other out.
What does the reaction make you want to do?
Sometimes the most revealing part of a countertransference response is not the feeling itself but what we find ourselves wanting to do.
If I feel protective, do I want to rescue? If I feel inadequate, do I want to prove myself? If I feel irritated, do I want to confront or leave? If I feel anxious, do I become more controlling or directive or appeasing? If I feel responsible, do I start doing more for the client than belongs within the therapeutic relationship?
The impulse itself is not necessarily problematic. Having an impulse is not the same as acting on it. A productive supervisory question might extend beyond ‘Why am I feeling this?’ to ‘What am I being drawn towards doing?’, and then, ‘What might happen if I did or didn’t do it?’
That can bring the relational pattern into view in a way that the original emotion alone may not. It’s also worth remembering that not every strong feeling in the therapy room is profound clinical data waiting to be decoded. Sometimes a difficult response is simply a proportionate response to what’s happening. Sometimes we’re tired or a situation is upsetting for straightforwardly human reasons. Part of what good supervision offers is discernment: help distinguishing a response that holds real clinical information from one that’s simply an ordinary human reaction, without needing to resolve that alone.
When the pattern travels into supervision
A pattern that develops between therapist and client can sometimes reappear in the supervisory relationship in a phenomenon usually called ‘parallel process’ (Hawkins and Shohet, 2012). A therapist who feels stuck with regard to a particular client may find themselves becoming stuck and unusually unforthcoming about that very client in supervision. A therapist who feels criticised by a client may notice themselves becoming uncharacteristically defensive with their supervisor regardless of what the supervisor has actually said or how they’ve said it.
This doesn’t mean that you (or your supervisor) are doing anything wrong. Rather, we can understand it as material that hasn’t yet been put into words finding its way into the relationship. The supervisor is then working with more than an account of the client. There is the client’s material, the therapist’s experience, the relationship between therapist and client, and what is happening between supervisee and supervisor as the material is discussed. This is one of the reasons why relational supervision can offer more than supervision that focuses predominantly on case content.
Why there’s rarely a clean answer
Some of the response may belong to me: my history, my sensitivities, my expectations, my current circumstances and the particular ways I have learnt to relate. Some of it may tell me something about the client: what they are communicating, how they relate to other people, what they may need, or how they experience relationships. And some of it may belong to the relationship itself: something that is being created, repeated or enacted between us. However it arises, my response is clinical information, not an objective measurement of the person sitting opposite me.
These are not mutually exclusive explanations. A client’s disappointment can touch my own fear of failing. My fear of failing can then lead me to work harder. My increased effort can alter the client’s experience of me. Their response to that can alter my response again. By the time we notice that something feels difficult, it may not be particularly useful to ask who started it.
What relational supervision can make possible
You don’t need to determine, in the moment, whether a strong reaction to a client began with you or with them. What you can do instead is notice the reaction, stay curious about what might belong to you without assuming it does, and keep asking whether your response is still in the best interest of the client or has drifted into being about your own. Bring it to supervision either way, including, and especially, when your first instinct is to keep it to yourself. That instinct is common and understandable, but it may be worth bringing to supervision precisely because it is making you want to stay silent.
The aim is not to become unaffected. It’s to have enough awareness and reflective space to notice when we’re reacting, enough curiosity to explore what the reaction might be telling us, and enough freedom to choose how we respond. When a client gets under your skin, the most useful question is therefore rarely, ‘Is this mine or theirs?’ It might be: ‘What’s happening here, and what do I need to understand before I decide what it means?’
Whether you’re a trainee or qualified therapist or a coach, if you’re looking for relational integrative supervision where you can think openly about your clinical or coaching work and your own responses within it, I’d be glad to talk. I offer a free 20 to 30-minute initial conversation to see whether working together might help, with no obligation to proceed beyond that. You can find more about my supervision practice, along with current fees, on my supervision page.
Garthine Walker is a UKCP-registered psychotherapist, clinical supervisor and EMCC-accredited coach based in Cardiff. She holds the Certificate in Clinical Supervision from the Metanoia Institute and offers relational integrative supervision to therapists and coaches online across the UK.
REFERENCES
Alina Elena Apostol, Kellie Turner, Rosa Hoshi and Aimee Pudduck (2025). ‘Contributory Factors to Self-Disclosure in Clinical Supervision: A Meta-ethnography’, Clinical Psychology & Psychotherapy, 32: e70068. DOI: 10.1002/cpp.70068.
Petrüska Clarkson (1990). ‘A Multiplicity of Psychotherapeutic Relationships’, British Journal of Psychotherapy, 7:2, 148-163. DOI: 10.1111/j.1752-0118.1990.tb01329.x
Richard G. Erskine (2023). ‘Presence and Involvement: Personal Perspectives on Countertransference’, International Journal of Integrative Psychotherapy, 14, 1-14. Available at https://www.integrative-journal.com/index.php/ijip/article/view/343.
Sigmund Freud (1910). ‘The Future Prospects of Psycho-Analytic Therapy’, in J. Strachey (ed. and trans.), The Standard Edition of the Complete Psychological Works of Sigmund Freud, Vol. 11. London: Hogarth Press, pp. 139-152.
Charles. J. Gelso and Jeffrey A. Hayes (2007). Countertransference and the Therapist’s Inner Experience: Perils and Possibilities. Lawrence Erlbaum Associates.
Peter Hawkins and Robin Shohet (2012). Supervision in the Helping Professions. Open University Press.
Paula Heimann (1950). ‘On Counter-Transference’, International Journal of Psychoanalysis, 31, 81-84.
Heinrich Racker (1957). ‘The Meanings and Uses of Countertransference’, Psychoanalytic Quarterly, 26:3, 303-357. DOI: 10.1080/21674086.1957.11926061.
Winnicott, D. W. (1949). ‘Hate in the Counter-Transference’, International Journal of Psychoanalysis, 30, 69-74.