The Four Attachment Styles: How Each Pattern Shapes Your Relationships

Where the model comes from

The four attachment styles describe patterns in how adults experience closeness, dependence, and trust. Their history begins with British psychiatrist John Bowlby, who proposed that bonds with caregivers influence expectations about safety and support. Mary Ainsworth studied infant-caregiver attachment through observed separations and reunions. Later researchers extended attachment theory to adult relationships, using different methods to study how adults think and feel about closeness.

In 1987, researchers Cindy Hazan and Phillip Shaver proposed that the same attachment system that binds an infant to a caregiver also operates in adult romantic love. They found that adults could reliably describe their approach to relationships in terms similar to the infant patterns, and that these self-descriptions predicted real differences in trust, jealousy, and comfort with closeness. That paper opened the field of adult attachment research that our test is built on.

Attachment style is not a personality type in the sense of a fixed label you are born with. It is a working pattern, learned early and reinforced over time, for what you expect from people you depend on and how you act when that dependency feels uncertain. It shows up most clearly in close relationships, romantic partnerships especially, but also in friendships, family, and even how people relate to a therapist.

The two dimensions behind the four styles

Modern attachment researchers, following the psychologists Kim Bartholomew and Leonard Horowitz, describe adult attachment along two continuous dimensions rather than four fixed compartments. The first is attachment anxiety: how much you worry about a partner's availability, how sensitive you are to signs of distance, and how strongly you need reassurance. The second is attachment avoidance: how uncomfortable you feel depending on someone or having them depend on you, and how strongly you value self-sufficiency.

Someone low on both dimensions tends toward secure attachment. Someone high on anxiety but low on avoidance tends toward anxious-preoccupied attachment. Someone low on anxiety but high on avoidance tends toward dismissive-avoidant attachment. Someone high on both leans fearful-avoidant, wanting closeness and fearing it at the same time. The developmental classification of disorganized attachment comes from a different assessment tradition and is not interchangeable with this self-report result.

Thinking in dimensions rather than boxes matters in practice. Most people are not a pure type; they sit somewhere on each dimension, and that position can differ from one relationship to the next. A person can look fairly secure with an old friend and much more anxious with a new romantic partner, because the dimension reflects a pattern of expectation, not a fixed trait stamped onto the person.

It is also worth understanding what the labels are not. Attachment style is not a personality test result in the sense of describing your general sociability or how organized you are, and it is not a clinical category. It is specifically about how you handle trust, closeness, and distance with the people you are closest to, which is why the same person can score quite differently on a broad personality inventory and on an attachment measure.

Attachment styles explained: dismissive avoidance and fearful avoidance

Dismissive-avoidant and fearful-avoidant descriptions both include discomfort with dependence, so pulling away alone does not distinguish them. In the four-style model, the difference is attachment anxiety. A dismissive description emphasizes keeping emotional distance without the same level of worry about abandonment. A fearful description combines distance with concern about whether closeness will be safe or dependable.

Consider what happens after a disagreement with a partner. Taking space might reflect a preference to handle feelings alone, or it might accompany a strong wish for reassurance that feels difficult to express. The outward action looks similar; the expectations differ. Ask about both closeness and worry rather than using one unanswered message to assign a style. Someone’s stated boundary is still a boundary, whatever explanation seems plausible.

The four styles at a glance

Secure attachment is a pattern of relatively low attachment anxiety and avoidance, associated with comfort seeking and offering support. People who lean secure are generally comfortable both depending on others and being depended on. They tend to trust that a partner will be there, handle conflict without catastrophizing, and recover from disagreements without long periods of doubt. Security is not the absence of relationship problems; it is a baseline of trust that makes those problems easier to work through together.

Anxious-preoccupied attachment centers on high attachment anxiety. People with this style tend to want a great deal of closeness, read a lot into small signs of distance, and feel a strong pull toward reassurance when a partner seems less available. The underlying fear is abandonment, and the underlying strength is often deep attentiveness and devotion to the people they love.

Dismissive-avoidant attachment centers on high attachment avoidance. People with this style tend to prize independence, feel most at ease when they are not too reliant on anyone, and pull back when a relationship asks for more emotional exposure than feels comfortable. The underlying strength is self-reliance and calm under pressure; the cost can be a partner feeling shut out.

Fearful-avoidant attachment combines high anxiety with high avoidance. People with this style often want closeness intensely and also find it frightening, which can produce a push and pull pattern of pursuing a partner and then withdrawing. An adult self-report result cannot establish that someone experienced childhood trauma or determine which therapy would help them. If closeness feels frightening, discuss the specific experiences and current difficulties with a qualified professional.

It is common for people to recognize a little of themselves in more than one description, especially before they have thought carefully about a specific relationship. That is expected, since the four labels are shorthand for positions on two continuous dimensions rather than airtight categories. The point of reading all four is less about finding a perfect match and more about noticing which pulls, toward closeness, toward distance, toward reassurance, feel most familiar in your own life.

How to find your attachment style

Our free Attachment Style Test is built on the two dimensions used by the Experiences in Close Relationships framework, the widely used measurement approach that grew out of Hazan and Shaver's original work. The items are written in plain, everyday language and ask about how you tend to feel and act in close relationships, particularly romantic ones, rather than how you wish you acted.

After you finish, the result screen names the style your answers lean toward, secure, anxious-preoccupied, dismissive-avoidant, or fearful-avoidant, and explains in plain language what that pattern tends to look like, where it often comes from, and what tends to help. If you want to go further, an optional in-depth report breaks your result down by the two underlying dimensions rather than a single label.

Use the Psychology.com Attachment Style Test at /tests/attachment-style with a particular close relationship in mind. Then use the healthy vs unhealthy relationships worksheet at /tools/healthy-vs-unhealthy-relationships to reflect on a concrete interaction. Write down what happened, what you expected, how you responded, and what you would like to ask for. Keep observations separate from guesses about the other person's motives.

If different attachment quizzes give you different labels, compare what they asked about before deciding that your pattern has changed. Shaver and Fraley's review of self-report measures explains that measures and classification methods can produce different category assignments. A quiz about a current partner and a questionnaire about close relationships generally may also invite different answers. Keep the relationship context alongside your result when you reflect on it.

Attachment styles over time: a changed score is not the whole story

Adult attachment can include both continuity and variation. Fraley and colleagues’ 2011 study followed attachment reports daily for a month in one sample and weekly for a year in another. The results favored a model with a stable component beneath temporary fluctuations. This supports looking beyond a single difficult week without assuming that adult attachment can never change.

If an attachment result changes, compare the relationship and circumstances you had in mind. Feeling uncertain during a new partnership and trusting a long-standing friend can coexist. A more useful record than a sequence of labels is what changed in asking for support, tolerating time apart, and discussing disagreement. Repeated experiences of those situations give the anxiety and avoidance dimensions context; a different category alone cannot explain the cause of the change.

What the research actually shows

Attachment theory rests on one of the deepest research bases in psychology, running from Bowlby's original observations through Ainsworth's structured studies of infants and caregivers, Hazan and Shaver's extension to adult romantic love, and Bartholomew and Horowitz's four-category model of adult styles. Decades of later work, summarized by researchers including Mario Mikulincer and Phillip Shaver, link attachment security with steadier relationships, better emotion regulation, and more resilient coping under stress.

At the same time, it is honest to say plainly what the research does and does not support. Adult attachment styles are best understood as two continuous dimensions, anxiety and avoidance, rather than four fixed boxes; the four-style language is a useful shorthand, not a diagnosis or a permanent identity. Attachment style can differ across relationships and can shift over a lifetime, as relationship experiences and circumstances change. Earned security is a term used in attachment interview research for secure adults who describe difficult early experiences; it is not a diagnosis or a guaranteed result of therapy.

The test on this site is a self-report screening tool for education and reflection. It is not a clinical instrument, and a result of anxious, avoidant, or fearful-avoidant is not a verdict on your character or your future in love. It is a description of a pattern you can understand, and in most cases work with and move, if you want to.

Common misconceptions

A common misconception is that attachment style is fixed from childhood and cannot be changed. Research allows for both continuity and change, and a current self-report score cannot reconstruct earlier caregiving. The cited earned-security study examined interview classifications and developmental histories; it did not test whether self-awareness, a partner, or a particular therapy reliably changes an online attachment result.

Another misconception is that anxious and avoidant attachment are opposite personalities that inevitably clash. These labels describe tendencies in relationships, and they do not establish who will be attracted to whom or whether a couple can work through conflict. Ask what happens during an actual disagreement: who seeks reassurance, who asks for space, and whether both people return to the conversation respectfully.

A third misconception treats attachment style as a label that explains everything about a relationship problem. It is one lens among several. Communication skills, life circumstances, values, and plain compatibility all matter too. Attachment style is most useful as a starting point for understanding your own reflexes, not as the final word on why a relationship is working or struggling.

A fourth misconception is that a secure result means you have nothing to work on and an insecure result means something is wrong with you. Neither is accurate. Security still benefits from active care, and an anxious, avoidant, or fearful-avoidant result describes a learned pattern with real strengths attached to it, not a deficiency to be ashamed of.

For example, if a delayed reply leads you to fear rejection, identify what you know about the situation before deciding what it means. If you need space during conflict, explain that need and agree on how to resume the discussion. These are conversation suggestions, not a treatment plan or a test of whether you are secure enough.

If these patterns cause persistent distress, the Psychology.com therapist directory at /therapist/ can help you look for professional support. You can bring examples of conflict, withdrawal, or reassurance seeking without having to choose a style first. Pressure, threats, and controlling behavior deserve attention in their own right; an attachment label does not excuse them.

When discussing a difficult interaction, describe the behavior you want to understand instead of assigning your partner a style. For example, "When we stop talking after a disagreement, I am unsure when to bring it up again" leaves room for an explanation and a request. Calling someone avoidant based on that interaction alone adds an interpretation that their own account may not support.