Why Do I Keep Attracting Narcissists?

RELATIONSHIPS & NARCISSISTIC ABUSE • July 21, 2026

The Question I Want to Answer Differently

Yolanda is 47 and sitting in the parking lot of her office building fifteen minutes before her first session with me, engine off, radio off, a paper cup of coffee going cold in the console. She’s a regional director for a healthcare logistics company, the kind of job where her calendar is color-coded in fifteen-minute increments and she has not, by her own count, taken a real lunch break since March. She has typed the same question into a search bar four separate times this year. Why do I keep attracting narcissists. She has read the quizzes. She has taken the quizzes. She is, she tells me in our first session, “pretty sure the problem is me.”

I want to say something plainly, in the first few minutes of this article, because I think it matters more than anything else I’ll say here: you are not a magnet. There is no clinical mechanism by which your nervous system, your childhood, or your attachment style causes another person to be controlling, exploitative, or cruel. In my work with driven women who ask this exact question, the honest answer is rarely the one they’re braced for. It isn’t that something is wrong with you. It’s that the question, as commonly asked, contains an assumption that doesn’t hold up: that repetition proves a hidden trait, and that the trait is yours to fix.

What actually explains why some people find themselves in more than one difficult relationship isn’t a single cause. It’s a combination of factors that have almost nothing to do with an internal “type” being recognized by someone dangerous. Selection pool. Timing. Isolation. The specific dynamics of dating apps, workplaces, and social circles. Financial pressure. Grief. Loneliness. Cultural scripts about what commitment is supposed to look like. Plain chance. None of these require you to have “attracted” anything. This piece walks through what the question usually means, what the research can and can’t tell us, how to tell a difficult personality from a diagnosable disorder from an abusive dynamic, and what a responsible next step looks like, including when that step needs to involve safety planning rather than self-reflection.

What “Narcissist” Actually Means, Clinically

“Narcissist” gets used in everyday language to describe almost anyone who was selfish, withholding, or hurtful. Clinically, it means something much narrower, and the gap between the popular use and the clinical use is where a lot of unnecessary self-blame gets manufactured.

I want to be direct about something I see often in my practice: driven women arrive with a fully formed diagnosis of their ex-partner, sometimes with more clinical vocabulary than some professionals use, and almost none of them are wrong that something painful happened. Many of them are working from an inaccurate label for what happened. That distinction matters clinically, and it matters for the reader’s own peace of mind, because chasing a diagnosis for someone else’s behavior can become a way of avoiding the more useful question: was this relationship safe, healthy, and mutual, regardless of what anyone was clinically diagnosable with.

There’s a reason this particular diagnostic language has become so available. Cultural fluency with narcissism as a concept has grown substantially over the past decade, and much of that growth has been genuinely useful: women who once had no vocabulary for what happened to them now have words like gaslighting, love bombing, and coercive control, and those words have helped many people name real harm they’d previously blamed themselves for. The trade-off is that the same vocabulary sometimes gets applied to situations it doesn’t accurately describe: a partner who was simply avoidant, a relationship that ended because two people wanted different things, a breakup that hurt without being abusive. Precision matters in both directions. Under-naming real abuse leaves women stuck defending someone who is actively harming them. Over-naming ordinary disappointment as personality-disordered abuse can leave a woman unable to trust her own read of future relationships, because she’s learned to see danger in situations that were simply not a match.

Narcissistic traits exist on a spectrum, and most people carry some of them without meeting criteria for a disorder. Selfishness during a hard season, defensiveness under criticism, a need to be right in an argument. These are common, sometimes exhausting, and not evidence of a personality disorder. What I look for clinically is not any single trait but a pattern: does this person’s behavior toward you and toward others remain consistent regardless of context, does it include a persistent absence of remorse, and does it involve deliberate exploitation of your trust rather than ordinary relational friction. That last piece, deliberate exploitation, is the hinge between a difficult relationship and an abusive one, and it’s the piece most search-engine quizzes skip entirely.

Traits, Incompatibility, Immaturity, or Abuse: Telling Them Apart

Here is a distinction I make constantly in session, because collapsing it is what keeps women stuck asking the wrong question. Not every difficult relationship is an abusive one, and conflating them does two kinds of harm: it can make a woman in a genuinely dangerous relationship minimize what’s happening because “he’s not a narcissist, he’s just stressed,” and it can make a woman in an ordinary incompatible relationship pathologize a partner who is simply not right for her, rather than recognizing an honest mismatch.

I recently sat with the framework developed by Bessel van der Kolk, MD, psychiatrist and trauma researcher, and found myself returning to a simpler distinction that I now use with almost every client asking this question: is what’s happening about incompatibility, immaturity, ordinary conflict, or is it about someone using coercion to gain and keep power over you. Those are different problems requiring different responses, and the language of “narcissist” often gets applied to all four without distinguishing them. This is also where a clear definition of gaslighting becomes useful, because gaslighting specifically describes an attack on your ability to trust your own perception, which is a different phenomenon than a partner simply being self-centered or difficult.

I first read Stark’s 2007 book on coercive control years into my clinical career, and it reorganized how I think about risk in relationships more than almost anything else I’ve studied since. What Stark’s framework clarifies for me clinically is that the presence or absence of a personality disorder diagnosis is not actually the most useful question when a woman is trying to understand whether her relationship is safe. The more useful question is behavioral: is your autonomy, your access to money, your relationships with other people, and your sense of your own perception being systematically restricted. That can happen with a diagnosable narcissist, with someone who has significant narcissistic traits but no diagnosis, with someone who has a different clinical picture entirely, including antisocial personality disorder, which involves its own distinct relationship to coercive control, or with someone who has no diagnosable condition at all and is simply choosing to behave this way. The behavior is what determines the danger, not the label.

On the other end of the spectrum, I want to name something I see just as often: relationships that are difficult, disappointing, even genuinely incompatible, that get retroactively relabeled as “narcissistic abuse” because that language is more available right now than “we were not well suited” or “he was emotionally immature and I deserved more consistency.” Both experiences are real. They require different next steps. A partner who is self-absorbed, avoidant, or bad at communicating under stress is a compatibility problem, sometimes a maturity problem, and it deserves honest evaluation rather than a diagnosis borrowed from a hashtag.

I find it useful to think about this as four separate categories that get flattened into one in casual conversation. There’s ordinary conflict: two decent people who disagree, get defensive, and need to repair, which is a normal feature of any relationship and not evidence of anyone’s character. There’s incompatibility: two people who are each fine on their own terms but don’t fit together, different values around money, family, ambition, or time, which is worth ending honestly rather than pathologizing. There’s immaturity: a partner who is self-centered, avoidant, or unable to sustain follow-through, often improvable with the right motivation and sometimes not, but not the same as deliberate exploitation. And there’s coercive control or abuse: a sustained, patterned effort to restrict another person’s autonomy and sense of self. Only the last category calls for the kind of safety planning this article discusses later. The first three call for honest conversation, couples work if both people want it, or an honest decision to part ways.

How This Actually Shows Up for Driven Women

Yolanda’s pattern, when we actually mapped it together over several sessions, did not look like what she’d expected. She had assumed we’d find some early wound that made her “pick” men who mistreated her. What we found instead was more mundane and, I think, more useful. Her first difficult relationship began at 29, when she was working eighty-hour weeks at a startup and her entire social world had narrowed to the six people on her team. Her second began at 41, three years into a demanding regional promotion, after a divorce that left her isolated from the friend group she’d shared with her ex-husband, meeting people almost exclusively through an app during a stretch when she had, by her own description, “no bandwidth to vet anyone properly.”

“I kept thinking there was a version of me who kept choosing this,” she said one Tuesday, turning a pen over in her hands. “But when I actually look at the timeline, both times I was completely underwater at work, completely cut off from the people who would have flagged something for me. Nobody in my life even met either of these men until we were already living together.”

Sitting with her, I felt something shift in how I understood the question she’d brought in. Not pity. Something closer to relief on her behalf. The spreadsheet she’d built in her head, the theory that some deep pattern in her was responsible, was not supported by her own data once she looked at it honestly.

What I’ve come to think of as the isolation-and-timing explanation accounts for far more repeated difficult relationships than any theory of unconscious repetition. Driven women in demanding careers often meet partners during exactly the windows when they have the least capacity to vet someone slowly: right after a big promotion, right after a loss, right after a move to a new city for work, right after a divorce when the existing friend group has fractured along old marital lines. None of that is a character flaw. It’s math. Less time with people who know you well enough to say something, more reliance on a stranger’s own account of themselves, more pressure to make a fast decision because loneliness has a cost too.

A few sessions later, Yolanda brought in something she’d noticed on her own, without my prompting, which is often when the most useful clinical shifts happen. “Both times,” she said, “I made the relationship the center of my life before I’d even had it checked out by anyone who loved me. And both times that wasn’t because something in me needed that. It was because I literally didn’t have anyone free for coffee that month.” I asked her what she thought would have been different with even one friend who’d met the person early. She sat with that for a while. “I think someone would have said something was off within a month. Both times. I just didn’t have that someone available.”

Deanna, 41, came to a similar realization from a different angle, and it took longer to get there because the first layer of her story looked, on the surface, like a punchline she’d already told herself. She runs a small private practice in Reno, and on the Tuesday she brought this up, she came in still wearing her lanyard from the office, the laminated badge with her own license number turned backward against her sweater. She’d had four sessions back to back and eaten a protein bar in the car between two of them. The wrapper was still in her coat pocket. She sat down, pulled her cardigan tighter, and said, “I need to tell you something and I need you to not make it about the fact that I do this for a living.”

“I have been telling myself for two years that I have a type,” she said. “That there’s some pattern in me, some radar I have for exactly the wrong man, because how else do you explain choosing the same disaster twice. And I have sat with clients who say this exact thing to me, word for word, and I have gently helped them see it isn’t true, and the entire time some part of me was thinking, well, it’s true for me. It has to be. I’m the one it kept happening to.” She laughed at that, but it wasn’t a real laugh. “A friend finally sat me down last month with both timelines written out. Same app. Same eighteen-month window. And I moved to this city twice in three years, once for a job that ended, once to be closer to my sister, and both times, both relationships started before I’d made a single friend who could have met either of these men and told me something was wrong. Nobody had eyes on it. Nobody had eyes on me.”

Sitting across from her, I felt the particular weight that shows up when someone who has spent years helping other people see their own blind spots finally turns that same clarity on herself, and doesn’t like what’s underneath it. It wasn’t relief, not yet. It was closer to grief, the kind that comes from realizing a story you told yourself for two years, a story that at least gave you something to work on, wasn’t the true one. I didn’t rush to correct her or reassure her. I let it sit there for a minute.

“It wasn’t a type,” she finally said, turning the empty protein bar wrapper over in her fingers like she’d just noticed it was there. “It was the only pool of people I had access to, twice, and nobody around me had the context to tell me anything, because I hadn’t let anyone close enough yet to have that context. That’s not the same as having a wound that keeps finding the same man.” She paused. “I think I liked the wound story better, actually. It made it mine to fix.”

Deanna’s version of the same isolation-and-timing pattern carried an extra layer worth naming honestly: being a clinician didn’t protect her from this, and it shouldn’t have been expected to. Clinical training teaches pattern recognition in other people’s lives, in a room with boundaries and time limits and someone else’s history already on the table. It doesn’t install a warning system for a person’s own dating life, conducted without a colleague present, during exactly the stretch when she had the least outside contact to check anything against. None of that made what happened her fault, and none of it means her judgment was compromised in some deeper, characterological way. It means she was a person, moving through two isolated years, same as anyone would have been.

She didn’t leave that session with a tidy answer. She left saying she wasn’t sure yet whether she believed the new version more than the old one, only that the new version had more evidence behind it. That’s where we left it.

This is one of the reasons I think the framework in a therapist’s guide to love bombing matters as much as it does. The tactic that makes an early relationship feel like fate is specifically effective when a person has less outside context to check it against, which is precisely the condition that isolation and time pressure create. It isn’t about a nervous system recognizing anything. It’s about the absence of friends who would have said, three weeks in, “this is moving unusually fast.”

Both/And: It Can Feel Familiar AND That Feeling Proves Nothing

Here is where I want to hold two things at once, because collapsing them into one or the other is where a lot of the “attracting narcissists” narrative goes wrong. It’s true that some women describe a difficult relationship as feeling intensely, even uncannily, familiar. And it’s true that this feeling of familiarity is not evidence of anything about the reader’s psychology, her childhood, or her worth. Both are true simultaneously, and the work is holding both rather than picking one.

The feeling of intensity or familiarity in a new relationship can come from a number of sources that have nothing to do with a person’s history repeating itself: a partner’s deliberate use of flattery and accelerated intimacy in the early weeks, which is specifically engineered to feel like uncommon compatibility; the simple chemistry of early romantic attraction, which produces intensity regardless of a partner’s character; grief or loneliness that makes attention from anyone feel significant; or, yes, sometimes a genuine echo of an earlier relationship, which still does not mean the echo caused the outcome. A feeling is data. It is not a verdict.

I think often of a caveat that Amir Levine, MD, psychiatrist at Columbia University, and Kenda Heller, MA, make explicitly in their book Attached, a book I recommend to clients more than almost any other on relationships: attachment style describes a general tendency in how a person tends to relate, not a fixed script that predicts or explains any specific outcome. An anxious or avoidant attachment style is one contributor among several to how a relationship unfolds. It is not a mechanism that causes someone to end up with a controlling partner, and treating it that way asks a concept designed to describe general patterns to do work it was never built to do.

What I say to clients who bring me this exact worry, that the intensity they felt somehow proves something damning about themselves, is this: the strength of an emotional pull is not a reliable measure of a relationship’s health, and treating it as one keeps women stuck in a loop of self-diagnosis instead of behavioral observation. The useful question is never “why did this feel so intense.” It’s “what did this person actually do, over time, when I asked for something, said no to something, or needed space.” That answer is observable. The feeling, on its own, tells you almost nothing. A more precise way to sort this out is outlined in distinguishing genuine red flags from your own trauma responses while dating, which walks through exactly this kind of behavioral observation rather than a feelings-based read.

The Systemic Lens: The Conditions That Shape Who You Meet

I want to widen the frame here, because an individual woman’s dating history doesn’t happen in a vacuum, and the terrain she’s dating within has changed in ways that make this question more common now than it might have been a generation ago. Modern dating operates through apps that reward the same self-presentation skills that make manipulation easier: a curated profile, a rehearsed opening message, an ability to perform compatibility before any real vetting has occurred. This is a structural condition of how people meet now, not a personal failing in anyone using the apps.

Financial precarity plays a role too, one I see constantly in the driven, high-earning women I work with, which surprises people who assume income insulates against this. A demanding career often means less unstructured time to build the kind of dense friend network that used to catch a concerning partner early. Long hours, frequent travel, relocation for a promotion. Each of these narrows the social circle exactly when a wider one would help most. None of this is about a woman’s internal wiring. It’s about the actual shape of the life she’s living and the pool of people that life gives her access to.

Cultural scripts add another layer. Women, and driven women in particular, are often taught to interpret persistence as devotion, jealousy as passion, and a partner’s intensity as evidence of how much they’re valued. Those scripts didn’t originate with any individual reader. They’re inherited, reinforced in film, in extended family expectations, in workplace culture that still rewards a woman for being endlessly accommodating. A woman raised inside those scripts isn’t broken for having absorbed them. She was taught a specific vocabulary for love that happens to overlap uncomfortably with some early tactics of control. If you want to understand where your own relational scripts were first written, a guide to the four attachment styles is a reasonable starting point, with the same caveat that applies throughout this piece: attachment style describes a general tendency, not a mechanism that causes anyone to attract harm.

Isolation itself, whether from a demanding job, a cross-country move, a divorce, caregiving responsibilities, or simply an adult life with a shrinking social radius, is one of the most consistent factors that researchers and clinicians who study coercive control point to as increasing risk of a harmful relationship taking hold, not because isolated women are more susceptible in some internal sense, but because isolation removes the outside perspective that usually interrupts a concerning pattern early. If you want to understand your own pattern accurately, the more productive audit isn’t of your psychology. It’s of your calendar, your circle, and your bandwidth during the months before each relationship began.

Workplace conditions deserve a specific mention because so many of the driven women I work with meet partners through their professional world, at conferences, in client relationships, through mutual colleagues, and workplace dynamics carry their own version of this risk. Professional settings reward polish, confidence, and a well-managed public presentation, exactly the qualities that make a manipulative person harder to distinguish from a genuinely impressive one in the early going. A woman evaluating a potential partner through the lens of professional competence, is he successful, articulate, well regarded by colleagues, is using a screening tool that was never designed to detect controlling behavior in a private relationship, because those two things are only loosely related. None of that is a failure of judgment. It’s a mismatch between the information a professional context provides and the information a safe relationship actually requires.

Chance plays a larger role than most of us want to admit, too. Some women date carefully, slowly, with a full circle of friends vetting every new partner, and still end up with someone who turns out to be controlling, because sophisticated manipulation is, by design, difficult to detect even under good conditions. Attributing every difficult relationship to a discoverable cause, whether that cause is framed as an internal pattern or an external condition, can itself become a way of avoiding an uncomfortable truth: sometimes a capable, careful person is still deceived, because deception is the entire point of the tactic being used against her.

Screening for Coercive Control, and What to Do if You’re Not Safe

Before anything else in this section: if you are afraid of a partner, if there has been physical violence, threats, strangulation, stalking, or a sense that leaving would be dangerous, please don’t wait to finish this article before getting support. In the United States, you can call or text 988 for the Suicide and Crisis Lifeline if you’re in crisis, call 911 for immediate danger, or contact the National Domestic Violence Hotline at 1-800-799-7233, available 24 hours a day, for confidential safety planning. None of what follows is a substitute for that kind of support if you need it right now.

Lenore Walker, EdD, ABPP, psychologist and professor emerita at Nova Southeastern University, is the researcher who first named what she called Battered Woman Syndrome in the late 1970s, describing patterns of psychological impact that can develop in a relationship marked by ongoing coercive control or violence. I want to be careful with how I introduce her work here, because it’s sometimes cited in ways that overreach: Walker’s framework describes one possible clinical picture among several that can result from sustained abuse, not a universal or inevitable outcome, and it was never intended as a tool for diagnosing whether a specific relationship is abusive. What her decades of research reinforce for me clinically is that the psychological effects of coercive control are real, measurable, and not a sign of personal weakness, which is worth naming plainly for any reader who has wondered why she didn’t leave sooner or see it clearly in the moment.

For readers trying to assess whether what they’re experiencing rises to coercive control, a few caveats matter clinically. Self-assessment tools, checklists, and articles like this one are educational, not diagnostic. They can help you organize your own observations, but they cannot replace an evaluation by a therapist trained in coercive control and intimate partner abuse, or a consultation with a domestic violence advocate, both of whom can assess your specific situation with more nuance than a general framework allows. If you find yourself checking most of the boxes on a coercive-control list, that’s a reason to seek a professional consultation, not a reason to confront your partner directly. Confrontation in a coercive-control dynamic can escalate risk rather than resolve it.

I want to be equally direct about what this article is not recommending. It is not recommending couples counseling as a first step if coercive control is present; that modality assumes a level of mutuality and safety that a controlling dynamic doesn’t have, and many trauma-informed clinicians will decline to do joint work in a case with active coercive control until individual safety has been assessed. It is not recommending disclosure to your partner that you’re evaluating the relationship this way. And it is not recommending that you leave, or stay, without an individualized safety plan built with a professional who knows your specific circumstances, including whether children, shared finances, or shared housing are involved. The guidance in preparing for a safe exit from a controlling relationship and in a guide to setting boundaries with a controlling family member both walk through what that kind of planning actually involves: documentation, financial separation steps, digital privacy, and a support network assembled before any conversation happens, not after.

Digital privacy deserves its own mention here, because it’s often overlooked. If you have any reason to believe your devices, accounts, or communications might be monitored, that changes how you should research any of this. Consider using a device your partner doesn’t have access to, and be aware that shared accounts, family sharing plans, and browser history can all be visible to someone with access to them. A domestic violence advocate can walk you through safer ways to gather information and make a plan without tipping off a partner who may escalate in response.

Finally, this article, and I as its author, are not able to tell you whether your specific partner has NPD, ASPD, or any other diagnosis, and that determination isn’t actually the thing that should drive your safety decisions. What should drive them is the behavior itself: are you free to see your friends, manage your own money, leave a conversation, say no, and trust your own memory of events. If the honest answer is no, that’s the signal that matters, independent of any label.

A Path Forward That Doesn’t Start With Blame

When Yolanda and I finished mapping her actual timeline, isolation and timing rather than a hidden pattern, something in her posture changed in session. Not relief exactly. She still had real grief to process about both relationships. But the shame that had been driving the search for “what’s wrong with me” loosened its grip enough for her to ask a more useful question: what would it take to build a life where she had people around her early enough to notice something concerning before she was eight months into it.

That’s the work I’d point most readers toward if this article resonates. Not a hunt for the hidden flaw that “attracts” harm, because there almost never is one, but a practical audit of the conditions that made vetting harder last time: How isolated was I. How much time pressure was I under. Who in my life actually met this person early, and did I listen to what they said. From there, the work is less about excavating a wound and more about building structural protections for next time: a friend who meets anyone new within the first month, a policy of not moving in together or making major joint decisions before a certain amount of time has passed, and, if there’s a genuine pattern of choosing incompatible partners rather than a coincidence of circumstance, working with a therapist on that specific pattern rather than assuming abuse explains it.

For readers who recognize a repeated pattern of choosing partners who aren’t right for them, separate from any abuse dynamic, Annie’s self-paced course Picking Better Partners is built specifically for that work; it isn’t open for enrollment yet, but you’re welcome to join the waitlist. For readers currently in an active coercive-control situation, the resources in the narcissism recovery hub are organized by stage, including the leaving and no-contact cluster, and are worth exploring before or alongside individual therapy.

Yolanda is still working on the practical side of this, months later. She’s rebuilt a monthly standing dinner with three women from an old job, a small thing, and told me recently that it feels less like a task and more like insurance. “I’m not doing it because I think I’m broken,” she said. “I’m doing it because I want three people who’ll actually meet the next guy before I’ve picked out furniture with him.” That’s not a cure for anything. It’s a structural fix for a structural problem, which is usually the more honest kind of progress therapy actually offers.

Of course this feels heavier than a quiz that promises a tidy answer. You’re not imagining how exhausting it is to have asked this question more than once. But the honest answer, most of the time, has nothing to do with a defect in you and everything to do with the ordinary, fixable conditions of an overextended life. That’s not a smaller story than the one about being unconsciously drawn to harm. It’s a more useful one, because unlike a hidden psychological flaw, isolation and timing are things you can actually change.

Warmly, Annie

Related Reading

Stark, Evan. Coercive Control: How Men Entrap Women in Personal Life. Oxford University Press, 2007.

Walker, Lenore E. The Battered Woman Syndrome, 4th ed. Springer Publishing Company, 2016.

Levine, Amir, and Kenda Heller. Attached: The New Science of Adult Attachment and How It Can Help You Find and Keep Love. TarcherPerigee, 2010.

Herman, Judith L. Trauma and Recovery: The Aftermath of Violence, from Domestic Abuse to Political Terror. Basic Books, 1992.

Medical Disclaimer