Attachment Style Self-Assessment (Printable)

How to score it honestly

Answer for how you have actually been across your last two or three significant relationships — not for how you are right now with one particular person, and not for how you would like to be. The most common distortion is answering as your best self in your calmest relationship, which returns secure for almost everybody.

Score each statement 0 to 4, where 0 is never true of me and 4 is almost always true. Add the six statements in each block. The blocks are anxiety, avoidance, disorganisation and security, and the highest total is the pattern that fits you most closely today.

Two totals within three points of each other is a genuine result, not a failure of the instrument. Attachment is dimensional, and mixed profiles are the norm rather than the exception. If anxiety and avoidance are both high, read the fearful-avoidant material.

Your result can change with the relationship you are in, and that is not a flaw in the questions. Attachment behaviour is relational: the same person can look anxious with an inconsistent partner and close to secure with a steady one. Re-score in six months and the movement will be informative.

Block A: Anxiety

How much of your attention the relationship takes when nothing is wrong.

These six ask about vigilance rather than about love. The construct is not how much you care — it is how much monitoring the caring requires: re-reading messages for tone, tracking response times, needing reassurance repeated because the last one has stopped working. A high score here is compatible with a very good relationship; what it measures is the cost of holding it, not its quality.

Block B: Avoidance

What happens as closeness increases.

The key word in every statement here is "as". Avoidance is not a dislike of relationships — it is a response that scales with proximity, which is why it so often appears three months in rather than at the start. These items ask about the direction of travel: whether interest reduces as access increases, whether you find reasons at the point things get settled, whether relief is a common feeling when plans fall through.

Block C: Disorganisation

Whether the two above take turns.

This block is not a third flavour of insecurity; it asks specifically about oscillation and about the conflict between the two systems. The distinguishing experience is wanting closeness intensely and finding it unbearable once you have it. The distinguishing behaviour is reversal — sending something and immediately wishing it unsent, ending something and immediately wanting it back. If Block C is high, read it alongside A and B rather than instead of them.

Block D: Security

Scored separately, because it is not simply low insecurity.

Security is its own set of behaviours rather than the absence of the other three, which is why it gets its own block. The items are about what happens under stress: whether you can say what you need without building a case first, whether conflict ends in repair rather than in a winner, whether an ordinary answer is enough to settle you. Someone can score moderately on A or B and still score high here — that combination is roughly what earned security looks like from the inside.

Reading your result

What the number is, and is not, evidence of.

The highest block is the pattern that currently fits you best. It is not a diagnosis — attachment style does not appear in the DSM-5 or ICD-11, and no label from a printed sheet should be treated as a finding about your character. Two things make the score more useful: writing down one specific incident that made you answer 4 on your highest-scoring item, and re-scoring in six months. An unexplained label predicts nothing; a label plus a dated incident is something you can work on.

For practitioners

This is free to print and use with clients, and it is deliberately not a validated instrument — where a measure is needed, the ECR-R or the AAI are the appropriate tools and this substitutes for neither. Its usefulness is as a conversation opener: the scoring is transparent enough that a client can see exactly which answers produced the result, which tends to move the discussion onto incidents rather than onto the label.

The scoring grid is on the last page and is designed to be photocopied on its own, so you can re-administer it later without reprinting the statements.