We’re an anxious and an avoidant couple
An anxious and avoidant pairing means one person’s worry about closeness pulls toward contact while the other’s discomfort with closeness pulls toward space. In a hard moment one reaches, the other steps back, and the stepping back brings more reaching. What matters most is whether the roles stay fixed, or move with the topic.
The first thing to look at is whether the roles hold across topics. If the person who reaches on one subject is the one who steps back on another, the loop belongs to the situation more than to either person’s style. If the roles never swap, and each person leans the same way with others, style may carry more of it.
If you feel afraid of them, or of what happens when you don’t go along with what they want, that matters more than any pattern on this page. Domestic abuse services talk to people who aren’t sure yet whether what’s happening counts.
If you or someone you know is having a difficult time, free support is available.
In an emergency, call your local emergency number.
Take the attachment style check together
Patterns that could explain it
Pursuer-distancer (demand-withdraw)
Fits if one presses for contact or a talk and the other steps back, and each move feeds the other.
One partner presses for contact or a talk and the other steps back, and each move brings more of the other. The research on the loop finds that the one who wants a change tends to be the one who presses, so the roles can switch with the topic.
Anxious-preoccupied attachment
Check whether one partner’s worry rises when the other seems far away, and settles only with contact.
One partner’s worry rises when the other seems far away or hard to read. It settles with contact or a reassuring word and can return within hours. Other close relationships may show the same worry.
Dismissive-avoidant attachment
Check whether the other partner handles strain alone and feels crowded when the first reaches.
The other partner handles strain alone and may go silent or get busy when feelings run high. Closeness can feel like pressure. It may show with friends and family as well, and not only with this partner.
Fits if the reaching grows louder as the answers get shorter: more messages, more questions, more checking.
The reaching grows louder as the answers get shorter: more messages, more questions, a check on when the other was last online. It can feel like the only way to be heard.
Fits if the other partner turns feelings down or turns away as closeness rises.
Feelings are turned down or put aside as closeness rises: the subject changed, a busy week, “it’s fine.” From outside it can look like indifference, though the body may still be reacting.
Fits if each partner carries one side of a shared fear and both stay in the arrangement though neither likes it.
Each partner may carry one side of a shared fear, one the fear of being left and the other the fear of being crowded. In a study of dating couples, some relationships that both partners rated fairly poorly were still together three years later, and the study doesn’t say why.
Understand it better
An example: Priya and Leo, the weekend after the argument
Priya and Leo have been together two years. On Thursday they argued about a trip Leo wants to take with his friends. They each describe what came next, separately.
Priya On Thursday I asked if we could talk about the trip. Leo said he was tired. I waited an hour and asked again. He said “later” and went to the gym. By Friday I had sent four messages and he had answered two, with one word each. On Saturday I wrote, fine, I’ll stop, and felt sick all day. Then he came over with dinner as if nothing had happened. I was so relieved I didn’t mention the trip.
Leo Thursday I was drained, and Priya wanted a long talk. I knew it would go in circles and end with me being the bad one, so I left. The messages after that read like a list of what I was failing at. By Saturday it was calm and I could think. I brought dinner because I wanted to see her. I didn’t bring up the trip because it was calm and I didn’t want to break that.
The trip is never discussed. Priya’s relief at the dinner and Leo’s wish to keep the calm end the round in the same way, which sets up the next one.
The weekend fits the pursuer-distancer loop: her asking brings his leaving, and his leaving brings more asking. Four messages by Friday fit hyperactivating strategies, where the reaching grows louder as the answers shrink. The gym and the calm by Saturday fit deactivating strategies, where feelings are turned down until they can be thought about. The research on this page asks whether that loop goes with a pairing of insecure partners. The relief at the dinner, on both sides, is the detail the page on collusion points to: each person’s fear is eased for a night, and nothing has changed.
Other details would point elsewhere. If Priya were the one who stepped back whenever Leo raised money, the roles would belong to the topic more than to either of them. If Leo were as short with his friends and his mother as he is with Priya, avoidant attachment would describe him better than the loop would. If the messages carried fear of what Leo might do, that would matter more than any pattern here. One weekend can’t settle which. What happens over several rounds, and on other subjects, says more.
What the research shows
Studies of this pairing are few, and most are questionnaires filled in by couples at one point in time. The research below looks at who gets paired with whom, at how paired attachment styles go with the push-pull loop, and at what the loop goes with.
Couples with an insecure partner reported more of the push-pull loop than secure couples, and couples with two insecure partners reported the most. The loop itself goes with poorer relationship and communication outcomes, whichever partner presses. In a long study of dating couples, partners were not paired at random and some poorly rated relationships lasted. None of the studies here follows an anxious partner and an avoidant partner through their arguments.
Among 354 dating couples, partners were not paired at random, and some poorly rated relationships lasted
In 1994, Lee Kirkpatrick and Keith Davis studied 354 heterosexual couples in serious dating relationships. Each partner reported an attachment style, and the couples were followed for three years.
Partners were not paired at random: no couple had two anxious partners and none had two avoidant partners. Relationships of avoidant men and of anxious women were rated fairly poorly by both partners at the start, yet they were surprisingly stable over the three years. That held after the researchers allowed for how long the couples had been together and how committed they were.
Doesn’t show Every couple was a man and a woman, and the findings differ by gender. The study counts who was paired with whom and who stayed together. It doesn’t say why, and it says nothing about how partners behaved in an argument.
Couples with an insecure partner reported more of the push-pull loop than secure couples
In 2008, Molly Millwood and Jennifer Waltz asked 147 couples to complete questionnaires on attachment style, on the gap between how much closeness each partner wanted, and on demand-withdraw communication. That is the loop in which one partner presses and the other steps back.
Compared with securely attached couples, couples in which one or both partners were insecurely attached reported bigger gaps in the closeness each wanted and more demand-withdraw. The authors note that, before their study, the loop had not been looked at through a couple’s specific pairing of attachment styles.
Doesn’t show Self-report at one point in time. The abstract groups couples as secure or as having one or two insecure partners, and doesn’t single out an anxious partner with an avoidant one. It can’t show that the pairing produces the loop, or which partner presses.
Couples with two insecure partners reported the most demand-withdraw
In 2009, Rachel Domingue and Debra Mollen studied 43 different-sex couples and 10 same-sex couples, all together at least two years and most living in one large US city in the Southwest. Both partners reported their attachment style and how they communicate during conflict. Couples were grouped as secure-secure, secure-insecure or insecure-insecure.
Secure-secure couples reported the most mutually constructive communication. Insecure-insecure couples reported the most demand-withdraw, and the most mutual avoidance and withholding.
Doesn’t show A small sample from one region, answered by the couples themselves at one time. The insecure-insecure group puts anxious, avoidant and fearful-avoidant partners together, so an anxious-avoidant pair isn’t told apart from two anxious or two avoidant partners. It can’t show which came first, the style or the communication.
Across 74 studies, the push-pull loop went with poorer outcomes whichever partner pressed
In 2014, Paul Schrodt, Paul Witt and Jenna Shimkowski pooled 74 studies with 14,255 people in all. Each looked at the demand-withdraw pattern, in which one partner presses for talk or change and the other pulls back.
The pattern went with poorer outcomes overall, a moderate link. It was about as strong when the wife pressed and the husband withdrew as when the husband pressed and the wife withdrew. The link was stronger in studies of distressed or clinical couples than in others, and stronger for relationship and communication outcomes than for well-being.
Doesn’t show The abstract says nothing about attachment styles. It can’t say whether the loop is happening in any one couple, why it starts, or that it causes the poorer outcomes.
Try it: the last time it went round
Take the last disagreement that went in circles and write a few lines about it. Keep the notes somewhere only you can open.
- Who reached first, and what the reaching sounded like
- What the other person did in the next hour
- How the round ended, and whether the topic was ever discussed
- A subject on which the roles have ever been the other way round
The last line is the one the research on the loop leans on. Roles that swap with the topic point to the loop more than to either person’s style. Roles that never swap, and that show with others as well, may point to something each person brings to close relationships.
Questions people ask
Can an anxious and avoidant relationship work?
The studies on this page don’t answer it. The one that followed couples over time reports its results by each partner’s style and gives none for anxious-avoidant pairs as such. What they show is narrower. In a study of dating couples, relationships of avoidant men and of anxious women, rated fairly poorly by both partners at the start, were surprisingly stable over three years. In questionnaire studies, couples with an insecure partner reported more of the push-pull loop than secure couples. And the loop goes with poorer relationship and communication outcomes across 74 studies. None of that says what any one couple will do.
Why do anxious and avoidant people attract each other?
The research on this page doesn’t show that they do. The nearest finding is from a study of 354 dating couples, where partners were not paired at random and no two anxious or two avoidant partners were found together. That counts who ended up with whom. It says nothing about who was drawn to whom, or why.
What is the anxious-avoidant trap?
A popular name for the loop in which one person reaches for closeness, the other steps back, and each move brings more of the other. It isn’t a research term. The research behind it is on the demand-withdraw pattern, which is the loop without the attachment labels, and on studies that ask whether couples with insecure partners report more of it.
Is the anxious partner always the one who pursues?
The studies here don’t say so. In the pooled studies the link with poorer outcomes was about as strong whichever partner pressed. The library’s page on the pursuer-distancer pattern adds that the pressing role tends to go to whoever wants something changed, so it can switch with the topic.
Which partner is to blame?
Nothing in this research assigns the loop to one partner. The studies measure it as something two people do together, each move following the other’s. The self-checks on this site say the same of themselves: a result is one person’s answers on the day, and neither says who is right or who should change.
Can two people take the attachment check together?
Each person takes the same short check, and both results appear side by side on one drawing, with a note on what the library’s research says about that pairing. A result is a person’s own answers on the day. It is not a type, a label or a diagnosis of either person.
Close to this
Go deeper
Other ways people say it
“anxious and avoidant relationship”
“I’m anxious and my partner is avoidant”
“my partner is avoidant and I’m anxious”
“anxious avoidant trap”
“can an anxious and avoidant relationship work”
“why do anxious and avoidant people attract each other”
“the more I reach, the more they pull away”
“anxious avoidant push pull cycle”
“anxious avoidant relationship cycle”
“we have opposite attachment styles”
Questions to carry
- In the last disagreement, who reached first, and what did the other person do in the next hour?
- Has there been a topic where we swapped places, with the one who usually reaches stepping back?
How this page was made. Built with Conscious Living AI and led by Andrew Fayad. How the library is made · Report an issue · Last changed November 11, 2026.