My Client Asked Me to Take On His Sister. Here's Why I Said No
What happens when clients want to refer a sibling, partner, or friend to the same therapist? The decision isn't only about confidentiality — it's also about how that familiarity can shape what happens in the therapy room
Doron had been in therapy with me for a few years, and finished about a year ago. He recently got in touch again — this time about his sister. He told me she'd hesitated a long time before agreeing to go to therapy at all, and what finally convinced her was what she'd heard about me from him (how much it had helped him, how much he trusted me).
"She only agreed to go to therapy because of you," he said. "There's no way I could have talked her into it otherwise." It was clear he meant this as a compliment, which made it especially hard to disappoint him when I said I couldn't take his sister on. He didn't understand the refusal. His own therapy was over, after all — so what exactly was the problem?
But a few years before that, a different client had told me about a colleague of hers from work who was looking for a therapist. After a few clarifying questions, I agreed. The colleague started her own process with me, and to this day it seems to be working well for both of them. My two answers were opposite each other, but it seems to me both were right. The question of how to decide whether to treat people who know each other keeps coming up, and it's never settled once and for all.
It's also a question the profession has no single agreed-upon answer to. Therapists from different approaches hold different criteria for when it's okay to take on someone who knows an existing client. What I'm describing here is a synthesis I've worked out for myself over the years, drawn from training, accumulated experience, a few mistakes I learned from, and various supervisors who guided me along the way. Another therapist, just as good, might have decided differently in those exact same two cases.
Clash of Versions
In any case, here's the rule that guides me: everything I know about a client needs to come to me from that client alone — not from a family member who isn't an integral part of the therapy, and not from Google. A large part of therapeutic work has to do not only with the content clients bring, but with how they choose to bring it.
What they emphasize, what they leave out, where the story they tell about themselves contradicts itself. And also, what I feel in the dynamic with them — which, to the best of my professional understanding, tells me more about them than about myself. That's the dynamic I work with, and any piece of information that reaches me from an outside source can damage my ability to understand it. For that same reason, by the way, I don't google my clients.
From this, maybe it's clearer why taking on someone who might, in their own therapy, talk about my client violates that condition. I want to stress this isn't only a question of confidentiality — which, with experience, I think becomes easier and easier to maintain.
The real issue is that it's enough for me to hold two versions of the same relationship in my head for my deep listening to each side to be compromised. I'm not an arbitrator. I want to see clients in their own terms.
Hoping I don't sound patronizing — a situation like this affects the clients themselves, even if it doesn't seem to: from the moment they know, even if it was at their own request, that their therapist also "knows" their sibling or their partner. That knowledge alone can make them censor themselves, even if no actual information was ever passed along.
Sibling War
Beyond that, over the years I've found that, for clients, my refusal to treat someone close to them doesn't always stay just a disappointment. Sometimes, when I hold a boundary like this, clients learn something from it about what happens when they themselves are in the room with me — for instance, that my conduct is dictated by commitment to their process, not just by a wish to please them. That's just one example, but the principle is that a refusal like this can actually strengthen mutual trust.
This question, incidentally, is well known to child psychologists. Parents who are very happy with one child's therapy often ask to bring a sibling to the same therapist. The request comes from the best possible place, of course, but those who treat children need to be especially careful.
They're already required to maintain constant loyalty to the child already in their care, and in their case that loyalty demands even greater vigilance than with those who treat adults — since the child assumes, rightly, that their parents may be exposed, to some degree, to the content of the sessions. When two children from the same family are with the same therapist, questions like who's loyal to whom, and what happens when one talks about the other, are bound only to heighten this tension — and might even derail the therapy.

When There's No Dependency Between Clients
One of the biggest differences between how clients think about this issue and how therapists think about it has to do with the time frame being considered. Like Doron, clients tend to think only about the present in this context. As far as Doron's concerned, his therapy with me is over, so there's no problem referring his sister to me.
But as therapists, we know that every ending of therapy is, on our end, a pause — in the sense that there's always the possibility that, a year or two from now, following a crisis or a change in life, a former client will want to come back, and we'll hold a place for them (not technically, but in principle — we'll go out of our way to find a way to meet with them again).
In Doron's case, suppose that during those years his sister had been in therapy with me, and then a fight broke out between them, or a rift formed. That would have put me in an impossible position. I couldn't be both of their therapist at the same time, and obviously I couldn't undo what had already happened. My refusal, then, is also an attempt to imagine a future I can't foresee.
By contrast, my agreeing to see the client's colleague came from the fact that the same check yielded a substantially different result. I found that the two of them like each other, but they're not part of each other's inner worlds. There's no power relationship between them at work, no dependency, and no reason for one to become a central figure in the other's therapy. The chance that they'll come up in each other's therapy in a way that requires real processing is low. And from my experience, when it's low enough, I can, in principle, agree.
A Crowded Social Circle
Family therapy, incidentally, is a completely different case, because there the structure is visible and known to everyone from the start. When a family comes to therapy, the client is the whole system — and then therapists can, and almost always should, see its members together in changing combinations — parents, siblings, everyone together — without that violating anything, because everyone knows they're part of the same process.
Even when a client in individual therapy asks to hold sessions with a partner, parent, or sibling, there's no problem, because it still serves one shared, agreed-upon process, and doesn't open a separate therapeutic space for that family member (even if I hold individual sessions with that family member too).
This principle wasn't always self-evident. When psychoanalysis was born in Vienna, Freud's own circle was itself a tight, closed social-professional circle. Analysts analyzed each other's family members, acquaintances, or partners, because their professional world was too small to allow full separation — and that was barely seen as a problem.
In that sense, no city in Israel is all that different from early-20th-century Vienna. In a small professional world, in limited territory, it's hard to keep circles fully separate. And because of that, even after all the considerations and checks, nothing is ever fully watertight. What looks like the right decision today may, in time, turn out to be a mistake. And that, too, is part of the work: to decide as best I can in the moment, and know that a day may come when I regret it.
Shall we check our calendars? If you would like to schedule a consultation, book me for a talk, or propose a collab, please send me an email at: amalia.rosenblum@gmail.com




