Why People Self-Sabotage Relationships — The Psychology
Last reviewed by the Men Women Psychology editorial team.
The evidence
What the research actually shows
Attachment research (Mikulincer and Shaver, 2007) suggests that people with avoidant tendencies often manage closeness by keeping partners at arm's length — minimizing dependence, deflecting intimacy, and withdrawing when a relationship deepens. People with anxious tendencies may sabotage differently, through reassurance-seeking, jealousy, or pre-emptive conflict driven by fear of abandonment. In both cases the behavior tends to be protective, not malicious.
Murray, Holmes and Collins's risk-regulation model (2006) frames intimacy as a constant trade-off between the rewards of connection and the danger of being hurt. People who feel less secure about their own worth, or about a partner's regard, tend to prioritize self-protection over closeness — pulling back at the first sign of risk, which can quietly erode the relationship they actually want.
Underlying much of this is the link between self-worth and how people behave in love. Those who doubt they are lovable, or who expect rejection, often act in ways that confirm those fears — distancing, criticizing, or withdrawing — in a self-fulfilling cycle that attachment researchers have documented across many studies. The dynamic shows up in both sexes, with individual differences far outweighing gender.
The mechanism
Why this happens
At the core is fear of vulnerability. Letting someone matter means giving them the power to hurt you, and for people whose early or past relationships taught them that closeness is unsafe, that prospect can feel genuinely threatening. Sabotage — picking a fight, finding flaws, going cold — can be an unconscious way to regain control and reduce that exposure.
The risk-regulation system helps explain the timing. Sabotage often intensifies precisely when things are going well, because deepening closeness raises the stakes of potential loss. Someone may withdraw or provoke conflict not despite the relationship being good, but because it has become good enough to be frightening to lose.
Low or contingent self-worth feeds the pattern. If a person believes they are not truly lovable, a partner's affection can create dissonance, and they may test, doubt, or undermine it to resolve the discomfort. Past betrayals and learned scripts about how relationships 'always end' add momentum, making the protective move feel like simple realism.
In practice
What this looks like in real life
Someone who finally meets a kind, available partner may suddenly fixate on small flaws or feel inexplicably bored — a common avoidant response when intimacy starts to feel too close for comfort.
A person with anxious tendencies might start arguments or issue ultimatums to 'find out' whether a partner will stay, inadvertently pushing away the very closeness they crave. The conflict feels like seeking reassurance; from the outside it can look like seeking the exit.
Someone who expects to be left may emotionally check out first — withholding affection or keeping options open — so that if it ends, they were 'never that invested.' Protecting against the pain of loss, they help bring the loss about.
Myth vs. evidence
What most people get wrong about this
The biggest misconception is that self-sabotage means a person does not care or is simply 'commitment-phobic' by nature. Research suggests the opposite is often true: the behavior tends to be strongest when someone cares a great deal, because that is when vulnerability — and the fear of losing it — runs highest.
It is also a mistake to see this as a fixed character flaw. Attachment patterns are learned and can shift. With awareness, security-building experiences, and sometimes therapy, people who tend to sabotage can gradually learn to tolerate closeness rather than flee it.
Why it matters
What this means for relationships
Recognizing sabotage as self-protection changes how partners can respond. Reacting to withdrawal with steady, non-punishing reassurance tends to help more than matching the distance, because it gently disconfirms the fear that closeness is dangerous. That said, the responsibility for change rests with the person doing the sabotaging, not only with the partner absorbing it.
For someone who notices the pattern in themselves, naming the fear in the moment — 'I'm pulling away because this matters' — can interrupt the automatic move. Building a more secure sense of self-worth, inside or outside the relationship, tends to reduce the pull toward self-protection over time.
Where it varies
The nuance
These patterns are not evenly distributed, but they are not strongly gendered either. Janet Hyde's gender similarities hypothesis (2005) reminds us that on most psychological measures men and women overlap far more than they differ, and self-sabotage is better predicted by attachment style and self-worth than by sex.
Not every relationship problem is self-sabotage. Sometimes pulling back is a healthy response to genuine incompatibility or mistreatment, not fear of intimacy. Distinguishing self-protective sabotage from sound judgment is part of the work, and honesty with oneself about which is which matters.
Questions people ask about this
What does it mean to self-sabotage a relationship?
It generally means acting in ways that undermine a relationship you want — withdrawing, picking fights, finding flaws, or keeping a partner at a distance. Research suggests it tends to be self-protective rather than deliberate, often driven by fear of intimacy, expected rejection, or doubts about one's own worth.
Why do people sabotage relationships when things are going well?
Risk-regulation research suggests that as closeness deepens, the stakes of potential loss rise. For someone who fears being hurt, a relationship becoming good enough to matter can feel threatening, and withdrawing or provoking conflict can be an unconscious way to reduce that exposure before they can be let down.
Is self-sabotage linked to attachment style?
Often, yes. Avoidant tendencies tend to show up as distancing and deflecting intimacy, while anxious tendencies may appear as testing, jealousy, or reassurance-seeking. These are patterns, not destiny — attachment styles are learned and can shift toward greater security with awareness and corrective experiences over time.
Do men and women self-sabotage differently?
There is overlap more than sharp division. On average, socialization can shape the form — for instance how withdrawal or conflict is expressed — but the underlying drivers of fear and low self-worth appear in both. Attachment style and personal history tend to predict the pattern better than gender does.
Can someone stop self-sabotaging?
Many people can, though it usually takes time. Noticing the pattern, naming the fear behind it, building a steadier sense of self-worth, and sometimes working with a therapist all tend to help. Steady, non-punishing reassurance from a partner can support change, but the core work belongs to the individual.
How can I tell self-sabotage apart from real incompatibility?
It can be genuinely hard. A useful clue is the pattern: if you repeatedly withdraw or provoke conflict specifically as things deepen, regardless of the partner, self-protection may be at play. If the concerns are specific, consistent, and about real mistreatment or mismatch, they may reflect sound judgment instead.
Research sources
These references point to the published research and established frameworks behind this page. They are provided for further reading; see our research methodology for how sources are selected.
- Mikulincer, M., & Shaver, P. R. (2007). Attachment in Adulthood: Structure, Dynamics, and Change. Guilford Press.
- Murray, S. L., Holmes, J. G., & Collins, N. L. (2006). Optimizing assurance: The risk regulation system in relationships. Psychological Bulletin, 132(5), 641–666.
- Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52(3), 511–524.
- Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults. Journal of Personality and Social Psychology, 61(2), 226–244.
- Hyde, J. S. (2005). The gender similarities hypothesis. American Psychologist, 60(6), 581–592.