Attachment Styles in Friendships: Anxiety, Avoidance, and Security Beyond Romance
Author: Ukrainian Psychological Hub · Published: September 22, 2026 · Editorial Policy
Attachment styles in friendships describe patterns in how adults experience closeness, availability, trust, support, and the possibility of rejection in close friendships. The strongest research model is dimensional: a person can be higher or lower in attachment anxiety and attachment avoidance within a particular friendship. Low anxiety plus low avoidance is commonly summarized as greater attachment security. Four familiar style labels—secure, anxious/preoccupied, dismissive-avoidant, and fearful-avoidant—can be useful shorthand, but they are not diagnoses, fixed personality types, or destinies.
For the broader taxonomy and overview of the four familiar patterns, see Attachment Styles in Adults. For the scientific model behind adult attachment, the anxiety × avoidance dimensions, internal working models, and measurement distinctions, see Adult Attachment Theory. This article focuses specifically on how those ideas apply to friendships beyond romance.
Friendship attachment also does not have to match romantic attachment. Research using relationship-specific measures shows that people can hold meaningfully different attachment representations with different close others, including parents, romantic partners, and friends. In a sample of more than 21,000 people, the Experiences in Close Relationships–Relationship Structures questionnaire was developed specifically to measure anxiety and avoidance across relationship contexts, and relationship-specific scores often predicted relationship outcomes better than broad attachment scores. Fraley et al. (2011).
That distinction matters for everyday questions such as “Why am I calm with my partner but anxious when my best friend does not reply?” or “Why can I be open with one friend and guarded with another?” Attachment research allows those patterns to coexist. It does not require one global label to explain every relationship.
What Does Attachment in Friendship Mean?
In adult attachment research, attachment is more specific than liking someone, spending a lot of time together, or calling someone a best friend. An attachment figure is a close other who may serve attachment-related functions such as being sought for comfort in distress, providing a sense of safe haven, supporting exploration as a secure base, and becoming someone whose absence or loss has particular emotional significance. Adult attachment networks are usually small and hierarchical rather than identical to a person’s entire social network. Gillath, Karantzas, and Lee (2019).
Friends can occupy important positions in those networks. A 2024 study of young adults found that friends remained highly important in attachment networks, even as romantic partners and parents often occupied different attachment functions. Friends were particularly prominent for some safe-haven functions, while romantic partners were especially important for separation-related functions. Carli et al. (2024). A large 2024 adult study likewise found that romantic partners were often primary attachment figures, while adults without romantic partners reported stronger emotional connection to friends and parents. Tian and Freeman (2024).
So the useful question is not simply “Can friends be attachment figures?” The answer is yes, they can. The more precise question is whether a particular friendship actually carries attachment functions for that person. A close friend may be central for emotional support without becoming the person’s primary attachment figure, and an ordinary friend may not function as an attachment figure at all.
Friendship Attachment Is Relationship-Specific
Adult attachment is sometimes discussed online as though each person owns one permanent style that appears everywhere. Relationship-specific research gives a more differentiated picture. General tendencies exist, but there is also substantial specificity across relationships. Klohnen et al. (2005) found that peer attachment models and parental attachment models show both organization and relationship-specific variation.
In a study of 512 emerging adults who rated attachment to a parent, romantic partner, and multiple friends, attachments across figures were correlated, yet much of the variance was specific to the attachment domain or individual relationship. Attachments to different friends were more similar to one another than to attachments in other domains, while still retaining substantial friend-specific variation. Bohn et al. (2023).
This means a person can plausibly feel secure with one friend, anxious with another, avoidant with a third, and relatively secure in a romantic relationship. It also means that a single “attachment style” quiz may flatten meaningful differences if it asks only about close relationships in general or only about romantic partners.
The Two Dimensions: Attachment Anxiety and Attachment Avoidance
Modern self-report research commonly organizes adult attachment around two continuous dimensions. Reviews of adult attachment measures distinguish this dimensional tradition from categorical systems and from interview-based attachment classifications. Ravitz et al. (2010).
Attachment anxiety in friendships
Attachment anxiety refers to the degree of worry about rejection, abandonment, insufficient care, or uncertain availability in a close relationship. In friendship, higher anxiety can make ambiguous social information feel especially important: a delayed reply, a changed plan, a new friend entering the group, a shorter message, or a period of lower contact may be interpreted through concern about losing the relationship.
Research does not support reducing anxious attachment to a simple stereotype of “needy people who always ask for reassurance.” Friendship findings are more complicated. In two dyadic studies of emerging-adult best friends, greater anxiety toward the best friend was associated with lower friendship quality, and one study found less self-disclosure and support-seeking among more anxious participants. Oldeman, Cillessen, and van den Berg (2025).
Other relationship-specific work shows why the picture can vary by context. In a 2024 study of 444 adults reporting on recent negative events and up to three attachment figures, greater relationship-specific anxiety predicted a greater likelihood of disclosing negative events, especially when perceived responsiveness was low; higher avoidance predicted a lower likelihood of disclosure. Snyder et al. (2024). The safest conclusion is that anxiety is associated with heightened concern about availability and rejection, while the behavioral expression of that concern depends on the relationship and situation.
Attachment avoidance in friendships
Attachment avoidance refers to discomfort with dependence, vulnerability, or emotional closeness and a stronger tendency to regulate distress through self-reliance or distance. In a friendship, higher avoidance may appear as reluctance to ask for help, difficulty disclosing vulnerable material, resistance to expectations of frequent contact, or pulling back when closeness feels demanding. Those are possible expressions of the dimension, not a checklist that identifies an “avoidant person.”
Friendship and social-network research links attachment insecurity to how people perceive, initiate, maintain, and dissolve social ties. Across several studies, Gillath and colleagues found associations between attachment insecurity and network tie strength, multiplexity, and network-management processes. Gillath, Karantzas, and Selcuk (2017). A later review summarized the broader literature as showing that insecurity is associated with less closeness and less effective network management, while emphasizing that attachment networks are structured and selective. Gillath, Karantzas, and Lee (2019).
Avoidance may also matter very early in friendship formation. In experimental and interaction studies, more avoidantly attached participants were especially likely to underestimate another person’s interest in becoming friends and showed lower interest partly through those perceptions. Beck, Aragón, and Clark (2025).
Low anxiety and low avoidance: attachment security
Greater security is usually represented by relatively low anxiety and low avoidance. In friendship, that configuration implies greater confidence that closeness is available without constant monitoring and greater comfort depending on another person without experiencing dependence as inherently threatening. Security does not mean perfect trust, constant contact, zero jealousy, or the absence of conflict. It describes a lower level of attachment-related anxiety and avoidance, not an idealized personality.
The Four Familiar Attachment Styles in Friendships
The four-style vocabulary comes from categorical and prototype traditions in adult attachment research. Bartholomew and Horowitz (1991) proposed a four-category adult model. Contemporary self-report work often treats the same territory dimensionally, which preserves information about how high or low a person is on anxiety and avoidance rather than forcing everyone into a box.
Secure attachment in friendship
A secure summary corresponds to relatively low anxiety and low avoidance. A person may be comfortable initiating contact, asking for support, giving a friend room, repairing after conflict, and tolerating ordinary fluctuations in contact without immediately interpreting them as abandonment or entrapment. These behaviors are examples of what lower anxiety and avoidance can permit; no single behavior proves that a friendship is secure.
Anxious or preoccupied attachment in friendship
An anxious/preoccupied summary corresponds to higher anxiety with relatively lower avoidance. The central theme is concern about the friend’s availability or the relationship’s stability. In daily life this may involve heightened attention to changes in tone, response time, invitations, closeness with other friends, or signs of exclusion. The pattern can coexist with a strong desire for intimacy. It should not be equated with being dramatic, dependent, jealous, or “too much,” because those labels collapse different motives and contexts into a personality judgment.
Dismissive-avoidant attachment in friendship
A dismissive-avoidant summary corresponds to relatively lower anxiety with higher avoidance. The central theme is distance from dependence and vulnerability. A person may value friendship while preferring less disclosure, fewer expectations of emotional availability, or more autonomy. That does not establish that the person lacks feelings or does not care. Attachment avoidance concerns how closeness and dependence are regulated, not whether someone possesses affection.
Fearful-avoidant attachment in friendship
A fearful-avoidant summary corresponds to higher anxiety and higher avoidance: closeness can be strongly desired while also feeling unsafe, exposing, or difficult to tolerate. That combination may create approach–withdraw dynamics within some friendships, but it should still be understood as a dimensional high-anxiety/high-avoidance pattern rather than a fixed type.
Fearful-avoidant attachment should not automatically be renamed “disorganized attachment.” Adult disorganization is measured in several different ways, including interview and self-report approaches, and the measurement literature remains heterogeneous. A 2023 systematic review identified 18 adult disorganization measures and concluded that the evidence base did not justify treating measurement as settled. Pollard, Bucci, and Berry (2023). Infant disorganized classifications, Adult Attachment Interview unresolved/disorganized classifications, adult self-report disorganization measures, and a high-anxiety/high-avoidance self-report pattern are related areas of research, not interchangeable labels.
What Attachment Can Look Like in Everyday Friendships
Attachment dimensions become especially visible when the relationship carries uncertainty, need, separation, vulnerability, or conflict. The same event can activate very different expectations depending on the friendship.
A friend replies much later than usual
Higher anxiety may increase attention to the delay and uncertainty about what it means. Higher avoidance may make it easier to disengage from the interaction or suppress the wish for reassurance. Greater security may support a wider range of interpretations while preserving the ability to ask directly when clarification matters. None of these reactions establishes an attachment style by itself; people also respond to actual history, stress, workload, conflict, culture, and communication norms.
A friend becomes close to someone else
For someone higher in attachment anxiety, a friend’s new relationship may become salient as a possible sign of replacement or declining importance. A person higher in avoidance may experience increased group closeness as pressure or may prefer less interdependence. A secure pattern does not eliminate jealousy or disappointment; it makes those emotions less automatically equivalent to evidence that the bond is ending.
You need emotional support
Attachment research is especially relevant to support seeking because attachment figures are defined partly by what happens under distress. Relationship-specific anxiety and avoidance can shape whether negative experiences are disclosed and how support is sought. Snyder et al. (2024). At the same time, friendship-specific work shows that anxious attachment does not produce one universal support-seeking pattern across studies. Oldeman et al. (2025).
There is conflict or a rupture
Attachment dimensions can influence whether a rupture is read as temporary, catastrophic, engulfing, or easier to avoid than address. Yet conflict patterns are dyadic. A 2025 study of 1,508 people organized into friendship groups found that judgments of friendship quality were strongly shaped by the unique relationship between two people, not simply by one person’s attachment-related characteristics. Avoidance showed a more consistent association with relationship-quality judgments than anxiety in that study.
Friendship Quality Is Dyadic, Not Just a Property of One Person
One of the most useful corrections to popular attachment discourse is that a friendship cannot be explained by attaching a label to one person. Two people create a history together. Responsiveness, reliability, reciprocity, communication, social context, competing demands, conflict, repair, and each friend’s expectations contribute to the relationship.
The social-relations evidence makes this concrete. In friendship quads, much of the variation in relationship-quality judgments came from the unique shared relationship between two friends. Dunnum and Chopik (2025). This is important because an anxious reaction may reflect a person’s attachment vulnerability, an actually inconsistent friend, or both. An avoidant reaction may reflect a stable preference for distance, a response to intrusive demands, or both. Attachment language becomes misleading when it erases what the other person is actually doing.
Can You Have Different Attachment Styles With Different Friends?
Yes. Relationship-specific attachment is one of the clearest findings relevant to friendships. The ECR-RS was designed precisely to assess attachment anxiety and avoidance across distinct relationship targets. Fraley et al. (2011). Research comparing parents, friends, and romantic partners also finds both cross-relationship consistency and substantial specificity. Bohn et al. (2023).
The same person might trust one long-term friend enough to disclose distress quickly, feel uncertain about another friend after repeated inconsistency, and keep a newer friend at emotional distance. Those differences do not automatically mean the person is inconsistent or “switching personalities.” They can reflect relationship-specific learning and expectations.
Can a Best Friend Become an Attachment Figure?
Yes, a best friend can function as an attachment figure when the relationship carries attachment functions such as safe haven and secure-base support. Adult attachment-network research documents friends as meaningful attachment figures, while also showing that attachment networks are hierarchical and that romantic partners or parents often occupy primary positions depending on age and life circumstances. Carli et al. (2024); Tian and Freeman (2024).
The functional definition also helps separate human friendship research from newer questions about nonhuman relationships. The separate article Can AI Become an Attachment Figure? examines whether some attachment functions can emerge in human–AI relationships. Evidence about AI attachment should not be used as evidence for how human friendships work, and human friendship studies should not be treated as proof of subjective experience in AI.
Attachment Style Is Not a Clinical Diagnosis
Secure, anxious/preoccupied, dismissive-avoidant, and fearful-avoidant are research or explanatory labels derived from attachment traditions. They are not DSM or ICD diagnoses. High attachment anxiety does not diagnose an anxiety disorder. High attachment avoidance does not diagnose a personality disorder. Fearful-avoidant attachment does not diagnose trauma or a disorganized attachment disorder.
Likewise, insecure adult attachment styles are not “attachment disorders.” Reactive Attachment Disorder and Disinhibited Social Engagement Disorder are separate clinical constructs with specific developmental and diagnostic criteria. Ordinary adult friendship patterns should not be renamed RAD or DSED.
Attachment Is Not the Only Explanation for Friendship Distress
Attachment is most useful when it clarifies expectations about closeness, availability, dependence, rejection, and support. It is less useful when it becomes a universal explanation for every painful social experience.
A delayed reply can matter because of attachment anxiety, but it can also matter because a friend has repeatedly disappeared without explanation. Wanting less contact can reflect attachment avoidance, but it can also reflect temperament, workload, burnout, neurodivergent social rhythms, a need for boundaries, or a friendship that no longer fits. Reassurance seeking can occur in attachment anxiety, but compulsive reassurance can also be part of obsessive-compulsive processes. The English Hub’s article on
Relationship OCD explains the obsession–compulsion cycle in relationship-focused OCD; that clinical process should not be reduced to an attachment label.
Attachment insecurity is associated with social anxiety in the broader adult literature, particularly attachment anxiety, but a systematic review found substantial heterogeneity and insufficient longitudinal evidence to infer causality. Manning et al. (2017). Measurement reviews also emphasize that adult attachment can be assessed in multiple ways that do not capture identical constructs. Ravitz et al. (2010).
How Are Attachment Patterns in Friendships Measured?
For friendship-specific questions, the measurement target matters. A questionnaire written only for romantic relationships should not automatically be treated as a direct measure of attachment to a particular friend. Researchers may use relationship-specific measures, broader adult self-report scales, interview methods, or other instruments depending on the construct.
The ECR-RS is especially relevant because it assesses attachment anxiety and avoidance in multiple relational contexts rather than assuming one undifferentiated style. Fraley et al. (2011). Reviews of adult attachment measurement caution that instruments differ in relationship focus, constructs, dimensional versus categorical scoring, administration, and psychometric support. Ravitz et al. (2010).
An online four-style quiz can be useful as a vocabulary prompt, but it cannot establish a diagnosis and may not tell you whether your answers describe a general tendency, romantic relationships, friendships as a domain, or one specific friendship. For a friendship question, that distinction is not technical trivia; it changes what is being measured.
What Does the Evidence Actually Show About Friendship Attachment?
The evidence base is meaningful but smaller and less uniform than the enormous popular discourse around “attachment styles” can make it seem. Several conclusions are reasonably well supported.
First, attachment can be relationship-specific. Large-scale measurement work and studies comparing attachment figures support meaningful within-person differences across relationships. Fraley et al. (2011); Bohn et al. (2023).
Second, friends can be part of adult attachment networks and can perform attachment functions, although they are not automatically the primary attachment figure. Carli et al. (2024); Tian and Freeman (2024).
Third, attachment anxiety and avoidance are associated with friendship processes such as perceived quality, support seeking, social-network management, and friendship initiation. The size and direction of particular behavioral associations can vary by outcome and context. Beck et al. (2025); Gillath et al. (2017); Oldeman et al. (2025); Snyder et al. (2024).
Fourth, friendship outcomes are strongly dyadic. Recent social-relations work shows that the unique relationship between two friends accounts for substantial variation in relationship-quality judgments. Dunnum and Chopik (2025).
The limitations are equally important. Much friendship-specific adult attachment research relies on self-report, cross-sectional or short-term designs, and samples of adolescents, students, or emerging adults. Measures and definitions differ across studies. Cultural norms around friendship, contact frequency, independence, family involvement, disclosure, and support can also shape behavior. These limitations make precise claims more defensible than sweeping statements such as “avoidants always have shallow friendships” or “anxious people always need constant reassurance.”
What Can You Do With This Information?
The most useful application is observation, not self-labeling. Instead of asking only “Which style am I?” ask relationship-specific questions: What happens in this friendship when I need support? What do I expect a delayed reply to mean? Do I feel able to ask directly for clarification? Do I suppress needs because dependence feels unsafe? Is my concern based mainly on uncertainty, or is the friend actually inconsistent? What happens after conflict?
It can also help to notice responsiveness without turning it into a scorecard. Feeling understood, valued, and cared for is an important relationship process. The Hub’s separate article on Perceived Partner Responsiveness reviews that construct in romantic relationships. The underlying evidence in that article is partner-focused, so it should not be silently generalized to every friendship; the broader idea is useful because attachment expectations interact with how available and responsive a specific relationship actually feels.
If a friendship repeatedly produces intense distress, compulsive checking, severe avoidance, inability to function, or patterns tied to trauma or another mental health condition, an attachment label is not a substitute for clinical assessment. Therapy can explore relational expectations and behavior, but there is no single standardized treatment called “healing your attachment style,” and no evidence-based promise that a person can be “reprogrammed” into a new type on a fixed timeline.
Frequently Asked Questions
Do attachment styles apply to friendships?
Yes. Adult attachment research includes close friendships, relationship-specific attachment measures, and adult attachment networks. The strongest formulation is not that every friendship has one of four fixed styles, but that anxiety and avoidance can be measured within friendship relationships and can differ across people and relationships.
Can I be securely attached in romance but anxiously attached to a friend?
Yes. Relationship-specific models can differ. The ECR-RS was designed to capture exactly this kind of variation, and empirical work finds substantial specificity across parents, partners, and friends. Fraley et al. (2011); Bohn et al. (2023).
What does anxious attachment look like in friendship?
It centers on heightened concern about rejection, abandonment, or uncertain availability. In some friendships that may involve sensitivity to delayed replies, exclusion, changes in closeness, or reassurance. Behavior is context-dependent, and research does not support a single universal checklist.
What does avoidant attachment look like in friendship?
It centers on discomfort with dependence or vulnerability and a stronger tendency toward distance or self-reliance. Possible expressions include reluctance to disclose distress, ask for help, or accept expectations of frequent emotional contact. These behaviors are not proof of an avoidant “type.”
Is fearful-avoidant attachment the same as disorganized attachment?
Not automatically. Fearful-avoidant is commonly used as a high-anxiety/high-avoidance adult self-report summary. Adult disorganization is a broader and methodologically heterogeneous research area, and infant disorganization and Adult Attachment Interview unresolved/disorganized classifications are separate assessment traditions. Pollard et al. (2023).
Can a best friend be my primary attachment figure?
Yes, for some adults a friend can become a primary or highly important attachment figure, particularly depending on romantic status, life stage, and the functions that relationship fulfills. Population-level studies still often find romantic partners or parents occupying primary positions, so friendship importance should be assessed rather than assumed.
Does anxious attachment mean I am too needy?
No. “Needy” is a social judgment, not an attachment measure. Attachment anxiety concerns worry about availability, rejection, and closeness. Whether a request for contact is excessive depends on context, agreements, actual responsiveness, and the needs of both friends.
Does avoidant attachment mean someone does not care about their friends?
No. Avoidance is about regulation of closeness, dependence, and vulnerability. A person can care deeply and still prefer more autonomy or find support seeking difficult.
Can friendships change attachment security?
Adult attachment representations are neither perfectly fixed nor infinitely fluid. Relationship-specific models can reflect ongoing relational experience, but this article does not treat one good friendship as a proven intervention or promise a specific trajectory of change. Questions about stability, change, and earned security belong to the cluster’s dedicated change articles when those pages are published.
Is attachment style in friendship a diagnosis?
No. Adult attachment style is a research construct. It is not a DSM or ICD diagnosis, and it should not be used to diagnose a friend from ordinary behavior.
What is the best way to measure attachment to a specific friend?
A relationship-specific measure is preferable when the question is specifically about one friend. The ECR-RS assesses anxiety and avoidance across relationship targets and has a large validation literature. Fraley et al. (2011). Choice of measure should still depend on the research or clinical purpose.
Related Articles
References
Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61(2), 226–244. https://doi.org/10.1037/0022-3514.61.2.226
Beck, L. A., Aragón, O. R., & Clark, M. S. (2025). Perceiving and responding to another’s interest in initiating friendship: The role of attachment insecurity. Journal of Social and Personal Relationships, 42(7). https://doi.org/10.1177/02654075251331796
Bohn, J., Holtmann, J., Luhmann, M., Koch, T., & Eid, M. (2023). Consistency and specificity of attachments to parents, friends, and romantic partners in emerging adulthood. Emerging Adulthood, 11(1), 58–73. https://doi.org/10.1177/21676968221081275
Carli, L. L., Alì, P. A., Anzelmo, E., Caprin, C., Crippa, F., Gallucci, M., Moioli, L., Traficante, D., & Feeney, J. A. (2024). Attachment networks in young adults. Frontiers in Psychology, 14, 1321185. https://doi.org/10.3389/fpsyg.2023.1321185
Dunnum, Z., & Chopik, W. J. (2025). A social relations perspective on attachment orientations and judgments of relationship quality in friendships. Journal of Research in Personality, 118, 104637. https://doi.org/10.1016/j.jrp.2025.104637
Fraley, R. C., Heffernan, M. E., Vicary, A. M., & Brumbaugh, C. C. (2011). The Experiences in Close Relationships–Relationship Structures questionnaire: A method for assessing attachment orientations across relationships. Psychological Assessment, 23(3), 615–625. https://doi.org/10.1037/a0022898
Gillath, O., Karantzas, G. C., & Lee, J. (2019). Attachment and social networks. Current Opinion in Psychology, 25, 21–25. https://doi.org/10.1016/j.copsyc.2018.02.010
Gillath, O., Karantzas, G. C., & Selcuk, E. (2017). A net of friends: Investigating friendship by integrating attachment theory and social network analysis. Personality and Social Psychology Bulletin, 43(11), 1546–1565. https://doi.org/10.1177/0146167217719731
Klohnen, E. C., Weller, J. A., Luo, S., & Choe, M. (2005). Organization and predictive power of general and relationship-specific attachment models: One for all, and all for one? Personality and Social Psychology Bulletin, 31(12), 1665–1682. https://doi.org/10.1177/0146167205278307
Manning, R. P. C., Dickson, J. M., Palmier-Claus, J., Cunliffe, A., & Taylor, P. J. (2017). A systematic review of adult attachment and social anxiety. Journal of Affective Disorders, 211, 44–59. https://doi.org/10.1016/j.jad.2016.12.020
Oldeman, M. G., Cillessen, A. H. N., & van den Berg, Y. H. M. (2025). Friendships in emerging adulthood: The role of parental and friendship attachment representations and intimacy. Personality and Social Psychology Bulletin, 51(4), 514–529. https://doi.org/10.1177/01461672231195339
Pollard, C., Bucci, S., & Berry, K. (2023). A systematic review of measures of adult disorganized attachment. British Journal of Clinical Psychology, 62(2), 329–355. https://doi.org/10.1111/bjc.12411
Ravitz, P., Maunder, R., Hunter, J., Sthankiya, B., & Lancee, W. (2010). Adult attachment measures: A 25-year review. Journal of Psychosomatic Research, 69(4), 419–432. https://doi.org/10.1016/j.jpsychores.2009.08.006
Snyder, M. D., Sun, E. R., Nie, J., & Jakubiak, B. K. (2024). When and how do people share? Attachment anxiety predicts support-seeking strategies. Personal Relationships, 31(4), 1113–1132. https://doi.org/10.1111/pere.12573
Tian, J., & Freeman, H. (2024). “All You Need Is Love” a social network approach to understanding attachment networks in adulthood. Behavioral Sciences, 14(8), 647. https://doi.org/10.3390/bs14080647