To Cross Or Not To Cross:
Do Boundaries In Therapy Protect Or Harm?

By Ofer Zur, Ph.D.

Online course on Boundaries in Psychotherapy for CE Credit

Source: Zur, O. To Cross or Not to Cross: Do boundaries in therapy protect or harm. Psychotherapy Bulletin, 39 (3), 27-32. Posted by permission of Division 29 (psychotherapy) of APA (updated 2018).

Original publication in PDF-printable format

Psychologists have been inundated with unequivocal messages about the depravity of boundary crossings and dual relationships in clinical practice. From graduate courses and texts on ethics, to continuing education workshops on “Risk Management”, to attorneys’ advice columns, we have been warned never to leave the office with a client, to be very careful about gifts, never to socialize with clients, to avoid bartering and to limit physical contact to a handshake or a pat on the back. We have also been cautioned that boundary crossings are likely to lead us down the slippery slope to exploitive sexual relationships. Boundary crossings and dual relationships have often been labeled unethical and often used synonymously with exploitation and harm.

This article will attempt to shed light on the complexities of boundary crossings and will clarify the relevant ethical and clinical concerns. It will distinguish between harmful boundary violations, beneficial boundary crossings and unavoidable or helpful dual relationships. Most importantly, it will suggest ways to increase clinical effectiveness by appropriately incorporating beneficial boundary crossing interventions into our clinical practices.

Defining Boundaries

Boundary issues mostly refer to the therapist’s self-disclosure, touch, exchange of gifts, bartering and fees, length and location of sessions and contact outside the office (Guthiel & Gabbard, 1993). Boundary crossing in psychotherapy is an elusive term and refers to any deviation from traditional analytic and risk management practices, i.e., the strict, ‘only in the office,’ emotionally distant forms of therapy (Lazarus & Zur, 2002). Dual relationships refer to situations where two or more connections exist between a therapist and a client. Examples of dual relationships are when a client is also a student, friend, employee or business associate of the therapist.

While most analysts, ethicists, attorneys and “experts” may use a broad brush in describing boundary issues, it is important that psychologists differentiate between harmful boundary violations and helpful boundary crossings. A boundary violation occurs when a therapist crosses the line of decency and integrity and misuses his/her power to exploit a client for the therapist’s own benefit. Boundary violations usually involve exploitive business or sexual relationships. Boundary violations are always unethical and are likely to be illegal. However, boundary crossings are often part of well-constructed treatment plans and, as such, they can increase therapeutic effectiveness (Lazarus & Zur, 2002). While all dual relationships involve boundary crossing, exploitative dual relationships are boundary violations. Obviously, not all boundary crossings are dual relationships.

Beneficial Boundary Crossings

While from the analytical point of view almost all boundary crossings are detrimental to the transference analysis and the clinical work, behavioral, cognitive-behavioral, Rational-Emotional, humanistic, existential, group, feminist, Ericksonian and family system psychotherapies often endorse many forms of helpful boundary crossings (Lazarus, 1994; Williams, 1997). Additionally, boundary crossing, when executed with the clients’ welfare in mind, is likely to enhance therapeutic alliance, the best predictor of therapeutic outcome.

Following are examples of beneficial boundary crossings and, when appropriate, the orientations or settings that support such interventions are included (Note: None of the following interventions constitute dual relationships):

  • Behavioral therapy endorses walking with an agoraphobic client to an open space outside the office or flying with a fear-of-flying client on an airplane as part of an exposure or in vivo intervention.
  • Child psychologists, and other psychologists who work with children, routinely leave the office for walks with them and or perhaps attend school plays in which they are performing. They also regularly touch and hug, provide snacks and drinks, play cards and exchange small gifts and photos with their young clients.
  • Albeit for different reasons, Cognitive, behavioral, cognitive-behavioral, feminist, group, humanistic, feminist and existential therapies all endorse self-disclosure as a way of modeling, offering an alternative perspective, exemplifying cognitive flexibility, creating authentic connections, increase therapeutic alliance or leveling the playing field.
  • Behavioral and family therapy support joining an anorexic or bulimic client for a lunch or for a family dinner.
  • Humanistic therapies are apt to frown upon therapists who never self-disclose, touch, hold, or hug their clients.
  • Many adolescent psychologists would not hesitate to go for a walk with a resisting, reluctant or irresponsive adolescent in order to break the ice. We have seen how this concept of boundary crossing has already filtered into our entertainment culture. Robin Williams, playing the therapist in the movie, Good Will Hunting, had the right attitude regarding boundaries when he decided to effectively break the ice by taking the highly resistive and distrustful young client, played by Matt Damon, to the riverbank for a walk.
  • Therapists who work with different cultures inevitably join their Native American clients in some of their sacred rituals, their Latin clients in weddings, their Catholic clients in confirmations, or their Jewish clients for Bar or Bat Mitzvahs. Refusing to do so in certain settings is likely to cause irreparable damage to the therapeutic alliance, nullify trust and render therapy ineffective.
  • Psychologists who work in poor, rural communities are often engaged in bartering arrangements, which may be the only way for people there to access mental health services. Bartering with cash-poor and art-rich artists is also a common practice.
  • If it is likely to benefit the client, therapists of many non-analytic orientations, will:
    • Go on a home visit to an ailing, bedridden or dying client. Such a visit also gives them a much better firsthand sense of the broader clinical context of their clients.
    • Take a depressed, medically non-compliant client on a vigorous walk.
    • Accompany a fearful client to a medically crucial but dreaded medical procedure.
    • Join a client-architect on a tour of her newest construction, a winery owner on a tour of his beloved winery or a proud sculptor to the opening of an exhibition of her work.
    • Escort a client to visit a gravesite or a place that held special meaning for the client and their deceased loved one in order to facilitate the grief process.
    • Join an addict at a first 12-step meeting.

Dual Relationships

Dual relationships are subtypes of boundary crossing. Psychologists practicing in rural and small communities encounter numerous unavoidable dual relationships in the course of their daily lives. The person who bags groceries in the supermarket, pumps gas, works in a dentist’s office or chaperones children on school field trips may often also be the therapist’s client. Relationships in such small communities can get even more complex when people choose their therapists because they know them and not because they saw their ad in the Yellow Pages. A therapist’s fellow congregation member, teammate in a local sports league or car dealer may all choose their psychologist because they have come to know him or her personally and they share values, attitudes, morals and or spiritual values. Like many other boundary crossings, such unavoidable dual relationships are not limited to rural or small communities; they are the norm within numerous small populations in larger metropolitan areas, e.g., gay/lesbian, handicapped, various minorities, religious congregations, and other such distinct small societies. In fact, duality, mutual dependence and prior knowledge of each other are prerequisites for the development of trust and respect in these communities. Non-sexual, non-exploitative dual relationships and familiarity between therapists and clients are not only normal but, in fact, increase trust. This enhances the therapeutic alliance, which is recognized as the best indicator of therapeutic results (Lambert, 1991; Norcross & Goldfried, 1992). Another excellent example is the military where, whether on a ship or in an isolated and remote base on foreign soil, dual relationships are not only unavoidable, but, in fact, mandatory.

It is important to differentiate between boundary crossing and dual relationships. Making a home visit to a bedridden patient or accompanying an acrophobic client to an open space, like many other ‘out-of-office’ experiences are boundary crossings that do not necessarily constitute dual relationships (Zur, 2001). Similarly, exchanging gifts, self-disclosure, bartering of goods (not services) or extending the therapeutic hour when needed are also boundary crossings but not dual relationships.

While dual relationships may be sometimes unavoidable, psychologists must nevertheless pay attention to the harm that can arise from them, especially where there is a conflict of interest. Conflicts of interest are often present in situations where the client is also a student, employee, employer or business partner. Of course, sexual dual relationships are always unethical, counter-clinical and illegal in most states.

The Ethics of Boundaries

Despite the prevalent belief to the contrary, there are no ethics codes or guidelines, which specifically deal with boundary crossings. The APA’s and almost all other professional organizations’ codes of ethics do not regulate non-sexual touch, gifts, length of sessions or self-disclosure. Of course, they all have a mandate to avoid harm and exploitation and respect clients’ integrity and autonomy. Back in 2002, APA Code of Ethics took a positive step regarding boundaries and dual relationships issues. It dropped the sentence, “Psychologists ordinarily refrain from bartering”, that appeared in the 1992 code and added the sentence, “Multiple relationships that would not reasonably be expected to cause impairment or risk exploitation or harm are not unethical” (APA, 2016, section 3.05), to the multiple relationships section (This view on multiple relationships also appears in APA Ethics Codes of 2010 and 2016). Just as important are clarifications that the new APA code provides in its Introduction and Applicability sections where it finally explains what some of the modifiers that are used in the Code (e.g., reasonably, appropriate, potentially) mean. More specifically it states: “As used in this Ethics Code, the term reasonable means the prevailing professional judgment of psychologists engaged in similar activities in similar circumstances, given the knowledge the psychologist had or should have had at the time” (APA, 2016). The importance of this clarification is that hopefully it will stop the experts, courts and ethics committees from using the analytic or urban yard stick to judge non-analytically oriented psychologists who strategically employ boundary crossing type interventions or work in small or rural communities where boundary crossing and dual relationships are unavoidable.

Boundary Crossings and the Standard of Care

The standard of care is defined as qualities and conditions that prevail or should prevail in a particular mental health service and that a reasonable and prudent practitioner follows. The standard is based on community and professional standards, as well as on state laws, case law, licensing boards’ regulations, a consensus of professionals, ethics codes of professional associations and a consensus in the community. The standard of care is not an objective yardstick to be found in any textbook. It is closely tied to a theoretical orientation. The examples of boundary crossings mentioned above clearly fall within the standard of care of behavioral, humanistic, family, and other non-analytic therapies. Regretfully, boards, courts and ethics committees too often confuse the standard of care with the analytic standards or with risk management guidelines (William, 1997). This confusion has caused tremendous injustice and immense suffering to therapists due to many boards’ and courts’ experts who routinely and mistakenly apply an analytic criterion and pronounce clinically appropriate boundary crossings and dual relationships, such as those mentioned above, to be below the standard of care.

The Slippery Slope Argument

There is a prevalent and unfounded belief in the ‘slippery slope’ argument, which claims that boundary crossings inevitably lead to boundary violations. It refers to the idea that failure to adhere to rigid boundaries and an emotionally distant form of therapy will ultimately foster exploitive, harmful and sexual dual relationships (Guthiel & Gabbard, 1993, Pope, 1990). This paranoid approach is based on the ‘snowball’ dynamic that asserts that giving a simple gift is the precursor of an exploitive business relationship; a therapist’s self-disclosure inevitably becomes an unhealthy social relationship; and a non-sexual hug will quickly devolve into a harmful sexual relationship. To allege that self-disclosure, a hug, a home visit, or accepting a gift are likely to lead to sex and harm is, in Lazarus’ words “an extreme form of syllogistic reasoning” (1994, p. 257).

Sexualizing Boundaries

The rigid attitude toward boundary crossings stems in part from what Dineen (1996) called ‘sexualizing boundaries.” This is a skewed view that sees all boundary crossings as sexual in nature as illustrated in the slippery slope argument. Simon (1991), for example, decrees that: “The boundary violation precursors of therapist-patient sex can be as psychologically damaging as the actual sexual involvement itself” (p. 614). Similarly, Pope (1990) states ” . . . non-sexual dual relationships, while not unethical and harmful per se, foster sexual dual relationships” (p. 688). These unreasonable beliefs link any deviation from risk management or analytic guidelines to sexual exploitation.

To Cross or Not to Cross

Intentional boundary crossings should be implemented with two things in mind: the welfare of the client and therapeutic effectiveness. Boundary crossing, like any other intervention, should be part of a well-constructed and clearly articulated treatment plan which takes into consideration the client’s problem, personality, situation, history, culture, etc. and the therapeutic setting and context. Boundary crossings with certain clients, such as those with Borderline Personality Disorders or those who are acutely paranoid are not usually recommended. Effective therapy with such clients often requires well-defined boundaries of time and space and a clearly structured therapeutic environment. Dual relationships, since they always entail boundary crossing, impose the same criteria on the therapist. Even when such relationships are unplanned and unavoidable, the welfare of the client and clinical effectiveness will always be the paramount concern.

Boundaries are like fences; they are man-made and are designed to separate. Their function is to “fence in” and “fence out”, to include and exclude. Being man-made, they can be constructed or dismantled, heightened or lowered, and made more or less permeable. Psychotherapy boundaries are an inherent part of the therapeutic setting. They have been the focus of psychoanalysts for clinical-transferential reasons. Consumer protection agencies, boards and professional organizations have focused on the boundary issue in order to guard clients from exploitative therapists.

The Concern with Rigid Boundaries

As to whether psychotherapy boundaries serve the protective purpose for which they were erected, I have two major concerns:

Firstly, I am concerned that rigid implementation of such boundaries decreases therapeutic effectiveness. As outlined above, there are numerous proven clinical and evidence-based interventions that fall under the heading of boundary crossings. These theoretically sound interventions are often not utilized due to therapists’ fears and their rigid adherence to risk management principles. As a result of this apprehension, many clients receive sub-standard care. Lazarus (1994) underscored that: “One of the worst professional or ethical violations is that of permitting current risk-management principles to take precedence over human interventions” (p. 260). Additionally, outcome research has documented the importance of rapport and warmth for effective therapy, and that rigidity, distance, and coldness are incompatible with healing. Appropriate boundary crossings and dual relationships are likely to increase familiarity, understanding, and connection hence, increasing clinical effectiveness (Lambert, 1991; Lazarus & Zur, 2002, Norcross & Goldfried, 1992).

Secondly, I am concerned that the isolation imposed by rigid boundaries increases the likelihood of exploitation of, and harm to, clients. Exploitation, as a rule, happens in isolation (i.e., child abuse, domestic violence, cults). As with any kind of abuse and exploitation, it is easier for predatory therapists to take advantage of their clients in the ‘darkness’ of isolation. In fact, many of our clients’ early life abuse and neglect was made possible due to the isolation of their families. The boundaries, which are supposed to protect clients from exploitation, also increase the therapists’ power and, therefore, increase the chance of a client being exploited (Zur, 2001).


Boundary crossing in psychotherapy has usually referred to any deviation from traditional analytic and risk management practices, i.e., strict, ‘only in the office,’ emotionally distant forms of therapy. They refer primarily to issues of self-disclosure, gifts, touch, bartering and home visits. Dual relationships, a sub-type of boundary crossing, refer to situations where multiple connections exist between a therapist and a client. Boundary crossings are different from harmful boundary violations and, appropriately employed, can increase clinical effectiveness and therapeutic outcome. Dual relationships and other forms of boundary crossing are unavoidable in many small and interdependent communities, such as rural, military, minority, church, university campus, and among gays, the deaf, etc. Unlike harmful boundary violations and sexual or exploitative dual relationships, neither boundary crossing nor dual relationships are unethical or below the standard of care. Behavioral, cognitive-behavioral, family, group, and existential therapeutic orientations are the most practiced orientations today. These approaches tend to endorse many types of boundary crossings that are considered clear boundary violations by many psychoanalysts and risk management advocates. In fact, feminist, humanistic, and existential orientations view the tearing down of artificial and rigid boundaries as essential for therapeutic effectiveness and healing. Boundary crossings should be implemented according to the client’s unique situation, condition, problems, personality, culture, and history and the setting in which therapy takes place. The rationale of boundary crossing, like any therapeutic intervention, should be articulated (in writing) in the treatment plan and consultations with experts are advised in complex cases. The unduly restrictive analytic risk-management emphasis on clearly defined, rigid, and inflexible boundaries often interferes with sound clinical judgment, which ought to be flexible and personally tailored to clients’ needs rather than to therapists’ dogmas or fears.


  • American Psychological Association (APA). (1992). Ethical Principles of Psychologists and Code of Conduct. American Psychologist, 47, 1597-1611.
  • American Psychological Association (APA). (2016). Ethical Principles of Psychologists and Code of Conduct. Retrieved from
  • Dineen, T. (1996). Manufacturing victims: What the psychology industry is doing to people. Toronto: Robert Davies Publishing.
  • Gutheil, T. G., & Gabbard, G. (1998). Misuses and misunderstandings of boundary theory in clinical and regulatory settings. American Journal of Psychiatry, 155 (3), 409-414.
  • Lambert, M. J. (1991). Introduction to psychotherapy research. In L. E. Beutler and M. Crago, Psychotherapy Research. Washington DC: American Psychological Association. pp. 1-11.
  • Lazarus, A. A. (1994). How certain boundaries and ethics diminish therapeutic effectiveness. Ethics and Behavior, 4, 253–261.
  • Lazarus, A. A. and Zur, O. (2002). Dual Relationships and Psychotherapy. New York: Springer.
  • Norcross, J. C. and M. R. Goldfried, M. R. (Eds.) (1992). Handbook of Psychotherapy Integration. New York: Basic Books.
  • Pope, K. S. (1990). Therapist-patient sexual contact: Clinical, legal, and ethical implications. In Margenau, E.A. The encyclopedia handbook of private practice. pp. 687-696. New York: Gardner Press, Inc.
  • Simon, R. I. (1991). Psychological injury caused by boundary violation precursors to therapist-patient sex. Psychiatric Annals, 21, 614-619.
  • Williams, M. H. (1997). Boundary violations: Do some contended standards of care fail to encompass commonplace procedures of humanistic, behavioral, and eclectic psychotherapies? Psychotherapy, 34 (3), 238-249.
  • Zur, O. (2001). Out-of-office experience: When crossing office boundaries and engaging in dual relationships are clinically beneficial and ethically sound. Independent Practitioner, 21 (1), 96-100.
  • Extensive Bibliography on Self-Disclosure in Psychotherapy and Counseling

Top of Page

Sign up for topical updates and invitations to participate with Dr. Zur