Discover your attachment style · Understand your relationship patterns
The Adult Attachment Scale (AAS) measures three subscales — closeness, dependence, and anxiety — mapping you to one of four attachment styles. Based on Bartholomew & Horowitz (1991). 18 questions, ~4 minutes.
Bartholomew (1991) four-category model · Click to read each style in depth
Comfortable with closeness and independence — positive views of self and others.
Craves closeness, fears abandonment — positive view of others, anxious view of self.
Values independence, deflects intimacy — positive view of self, distancing view of others.
Wants closeness but fears it — negative view of both self and others.
The Adult Attachment Scale (AAS) is a self-report measure of how adults relate in close, emotionally intimate relationships. Developed by psychologists Nancy Collins and Stephen Read in 1990, it contains 18 items spanning three subscales: comfort with closeness, dependence on others, and anxiety about abandonment.
We map results onto the Bartholomew & Horowitz (1991) four-category model — secure, anxious-preoccupied, dismissive-avoidant, and fearful-avoidant — a framework widely used in research on intimate relationships, couple conflict, parenting, and emotion regulation.
Attachment measures have a research basis, but they are *not* personality diagnoses or fate predictions. Your result reflects how you currently tend to relate in close relationships — a useful starting point for self-awareness, not a final verdict on who you are.
Each answer is scored from 1 to 5. Items 2, 7, 8, 13, 16, 17, and 18 are reverse-scored (reversed score = 6 − original score). We then calculate the mean of six items for each subscale: Close, Depend, and Anxiety.
For the four-style result, we average the Close and Depend means, then use 3 as the boundary for that combined score and for Anxiety. High Close/Depend with low Anxiety maps to Secure; both high maps to Anxious-Preoccupied; both low maps to Dismissive-Avoidant; low Close/Depend with high Anxiety maps to Fearful-Avoidant. A score of exactly 3 is placed in the high group.
This is this site's implementation for mapping the three continuous Collins & Read subscales to four attachment styles. Read the three continuous scores alongside the label; the category is not a diagnosis or a fixed identity.
The four styles differ systematically along two dimensions: your baseline expectations of your own worth, and your baseline expectations of whether others are reliable. The table below offers a quick comparison; click any row to read the full description.
| Attachment style | View of self | View of others | Common relationship pattern |
|---|---|---|---|
| Secure | Generally positive | Generally positive | Comfortable with closeness and autonomy; communicates needs directly |
| Anxious-Preoccupied | Negative; fears unworthiness | Idealised; fears abandonment | Craves closeness, hyper-vigilant for rejection signs |
| Dismissive-Avoidant | Self-reliant, positive | Distrustful; values distance | Suppresses attachment needs; deflects intimacy |
| Fearful-Avoidant (Disorganized) | Negative | Negative; mistrustful | Wants closeness yet fears it; oscillates between approach and withdrawal |
Attachment theory originated in the work of British psychoanalyst John Bowlby in the 1950s–60s, who proposed that infants form "internal working models" with primary caregivers that shape later relationship expectations. Mary Ainsworth's "Strange Situation" experiments classified infant attachment into secure, anxious, and avoidant patterns.
In 1987, Cindy Hazan and Phillip Shaver extended attachment to adult romantic relationships with a three-category model. Collins and Read (1990) then published the AAS, measuring attachment along three continuous subscales. Bartholomew & Horowitz (1991) introduced a two-dimensional framework — view of self × view of others — yielding the four-category model used here.
Subsequent research has replicated these constructs across multiple cultural contexts, with attachment correlates (relationship satisfaction, conflict resolution, emotion regulation) supported in many samples. Researchers also note that the specific expression of attachment may vary by culture, so cross-cultural use warrants caution.
Yes. Attachment styles are patterns, not fixed traits. Self-awareness, secure relationship experiences, and therapy can shift insecure styles toward "earned secure" attachment. Research suggests meaningful change is possible across years, though the patterns formed in early life often resurface under stress.
The Adult Attachment Scale (Collins & Read, 1990) is one of several validated self-report measures of attachment. It has been used in hundreds of peer-reviewed studies. That said, all self-report measures have limits — your result reflects how you tend to think about close relationships, not a fixed verdict.
18 questions, 5-point Likert scale. Most people finish in 3–5 minutes. There are no trick questions; first instinct is usually most accurate.
No. This is a self-reflection tool, not a diagnostic instrument. Attachment style is a research construct, not a medical diagnosis. If you have ongoing relationship distress, persistent emotional pain, or trauma symptoms, please consult a licensed mental health professional.
No. The test runs entirely in your browser. Your answers are not sent to a server, not stored in any database, and not linked to your identity. Closing the page erases the result.
Most people show some traits from multiple styles — categorical labels are simplifications of an underlying continuum. Your assigned type reflects your dominant pattern. Reading the descriptions of adjacent types often gives a fuller picture.
Attachment style measures how you behave in close emotional relationships, not your overall personality. MBTI and similar tests assess broader traits like introversion or thinking style. Attachment is more behavioral and shaped by early caregiver experiences.
Insecure attachment is not a character flaw — it usually reflects what worked to stay safe in early relationships. Helpful steps: read about your style in depth, notice triggers in current relationships, communicate needs explicitly, and consider attachment-informed therapy if patterns cause distress.
Privacy: All calculations run entirely in your browser. Answers are never uploaded to a server, stored in any database, or linked to your identity. Closing the page erases the result.
Not a diagnosis: This test is for educational and self-reflection purposes only. It is not a clinical diagnosis and is not a substitute for professional mental health consultation or therapy. Attachment style is a research construct, not a medical diagnosis.
If you are experiencing any of the following, please seek professional support:
Crisis resources: US: 988 Suicide & Crisis Lifeline (call or text 988); National Domestic Violence Hotline 1-800-799-7233. International: findahelpline.com. If life is in immediate danger, call your local emergency number or go to the nearest emergency room.
18 questions, about 4 minutes. No signup. Answers stay in your browser.
Start the test