Why do I lose interest as soon as someone likes me back?
The short answer
The best supported explanation is attachment avoidance. For some people closeness is itself the thing that sets off withdrawal, and someone liking you back is the precise moment the relationship stops being hypothetical. A second and much weaker strand suggests uncertainty about whether someone wants you can hold attention, so resolving it removes something that was doing work. Neither makes this a diagnosis, and neither says anything about what you owe the person once the feeling drops.
What readers think
When the feeling dropped, what did you actually do about it?
Anonymous. One vote per reader.
It has a shape people recognise long before they can explain it. The interest is real, sometimes consuming, and it survives weeks of not knowing. Then a message arrives saying the feeling is returned, and within days something has gone flat that had been vivid the week before. Most people explain this to themselves as evidence that they are shallow, or broken, or built wrong. The research points somewhere less dramatic and considerably more useful: a description of what proximity does to some nervous systems, and a smaller, shakier finding about what the not knowing had been contributing all along.
The strongest account is about closeness, not about them
Attachment researchers describe a set of responses called deactivating strategies: when the attachment system is roused by closeness or dependence, some people manage the arousal by turning it down rather than by seeking contact. In practice that looks like minimising the importance of the bond, becoming suddenly alert to the other person's flaws, redirecting attention elsewhere, and withdrawing[1]. The timing matters here more than anything else. While the other person's feelings are unknown, nothing is being asked of you and there is nothing to withdraw from. Reciprocation converts an idea into a situation, and it is the situation, not the person, that the response is aimed at. A daily diary study of undergraduates in New Zealand caught this at close range: on days when more avoidant participants felt they had less personal control in the relationship or felt their partner's influence more strongly, they reported lower perceived regard, more derogation of the partner, and more withdrawal that same day[2]. That is this whole pattern compressed into twenty four hours.
Why it can feel like the attraction itself dies
The experience people describe is rarely a decision to pull back. It is that the wanting simply stops, which is why it gets read as proof that the feeling was fake. There is evidence that the negative charge reaches into the parts of the relationship people assume are immune. A study combining a large self-report sample with a forty two day diary of couples found that attachment avoidance was associated with more aversive feelings and thoughts during sex, while attachment anxiety was associated with a more ambivalent construal of it[3]. Sex is usually treated as the one place where closeness is uncomplicatedly wanted, so a finding that avoidance colours it too is worth sitting with. It suggests the response is not a judgement about whether this person is good enough. It is triggered by the closeness itself, which is exactly why it arrives at the moment the closeness becomes available and why it can feel indistinguishable from the attraction having ended.
What the uncertainty finding actually shows, and what it does not
There is a well travelled study in which college women viewed profiles of men who had supposedly already seen theirs. Told the men liked them a lot, liked them an average amount, or that it was one or the other, the women in the uncertain condition reported the most attraction, more even than those told they were liked a lot[4]. Online this has hardened into a law about human nature. The follow up work does not support that. A later set of studies separated liking from the wish to interact and failed to reproduce the original result: uncertainty raised the desire to meet and talk with someone, but not liking, and not sexual interest, and in a within participant design people became more attracted once they knew they were liked[5]. The original had forty seven participants, all women, all undergraduates, looking at profiles of strangers they would never meet, which meant there was no possibility of being rejected. What survives is narrow and still worth having: not knowing can keep attention fixed on someone, and attention is easily mistaken for desire. That is a real thing to have lost when the answer arrives. It is not the same as wanting what you cannot have.
The usual direction is the opposite, which is the informative part
Reciprocity of liking is among the sturdier effects in attraction research. Learning that someone likes you tends to make you like them more, and the process runs through several channels at once, affective, cognitive and behavioural, with the perception of good intentions doing much of the work[6]. There is an important qualifier. In speed dating research, unique desire from one particular person prompted unique desire in return, but people who tended to want everyone they met were wanted less in return[7]. Two things follow. The first is that if being liked back reliably cools you, you are running against the modal finding rather than illustrating it, which is genuinely informative: it points at your own history rather than at how attraction works in general. The second is that the qualifier offers one ordinary alternative explanation. If someone's interest reads as undiscriminating, as something they would have offered anyone, the drop may be information about what their interest told you rather than a reflex against closeness.
A dimension, not a diagnosis, and not a permission slip
Avoidance is not a condition anyone has. When researchers examined how people actually answer adult attachment questionnaires, the responses spread along continuous dimensions rather than falling into types[8], so the honest statement is that you score somewhere on a scale that everyone is on, using an instrument that asks you to report on yourself and inherits every limitation that implies. Nor is the position fixed. Two intensive longitudinal studies, one tracking people daily for a month and another weekly for a year, found the data fitted a model with a stable underlying factor plus real fluctuation on top of it[9]. Stable enough to be a pattern, loose enough to move. That combination is also where the ethical line sits. Recognising a pattern and adopting it as an identity are different acts, and the difference shows in what happens in the fortnight after the feeling drops: whether you say so plainly or go quiet, whether you keep accepting someone's attention while giving nothing back, whether the explanation gets used to understand yourself or offered to the other person as though it settled their complaint. Nothing in this literature bears on that. It explains why the feeling changed. It says nothing about what you owe someone once it has.
It depends, on what?
What changes the answer
There are plainer explanations that should be ruled out before reaching for this one. Interest sustained across a long stretch of not knowing is often interest in a version of someone you assembled yourself, and that version cannot survive contact with the actual person; when it collapses, nothing has been triggered, a projection has simply ended. Sometimes it really is the individual, and the pursuit was momentum rather than attraction. Sometimes the drop is accurate. The avoidance account also has a failure mode worth naming: it can explain any ending whatsoever, and a story that cannot be wrong is not doing much work. Treat it as the leading candidate when the pattern has repeated across quite different people, and as a weak guess after one instance. The sampling limits are real too. The daily diary work here involved seventy four New Zealand undergraduates, the uncertainty study forty seven American college women, and most of this field is built on young, educated, Western samples. That overlaps with a good part of who asks this question and describes no one else particularly well.
Where people disagree
Both sides of it
Reasonable people land in different places on this. Here is the strongest version of each case, not a straw man of the one we disagree with.
The case for yes
- The timing fits the avoidance account precisely, since reciprocation is the exact point at which closeness stops being theoretical and starts making demands.
- Daily diary evidence shows the withdrawal happening within a single day of increased dependence, which is the same sequence at a smaller scale.
- The account explains why the change feels like the attraction dying rather than like a decision, because deactivation operates below the level of choice.
- It predicts something checkable: if closeness is the trigger, the drop should recur across quite different people rather than tracking anything about them.
The case for no
- Avoidance can be stretched to explain any ending, and an account that no outcome could contradict is close to useless.
- Attachment measures are self-report, so someone reading about avoidance and recognising themselves is not the same as having been assessed.
- The uncertainty half of the popular explanation rests on a small study that later work did not reproduce cleanly, so it should carry very little weight.
- Ordinary explanations frequently fit better, including that the person was more interesting as a prospect than as company, which requires no theory.
Evidence
What the research says
- Moderate evidenceTheoretical review across many studies
Attachment theory describes deactivating strategies, in which closeness or dependence prompts people high in avoidance to suppress attachment needs, minimise the relationship's importance and withdraw, rather than to seek contact.
Mikulincer, M., Shaver, P. R., & Pereg, D., Attachment Theory and Affect Regulation: The Dynamics, Development, and Cognitive Consequences of Attachment-Related Strategies (2003)(opens in a new tab)What this does not show
This is a theoretical synthesis of many studies rather than a single test, and the underlying evidence is mostly correlational and cross-sectional. It describes a regulatory tendency at group level and does not establish that any individual's withdrawal has this cause.
- Moderate evidence74 undergraduates reporting on daily interactions
In a daily diary study, on days when more avoidant participants perceived less personal control or greater partner influence, they reported lower perceived regard and intimacy, greater derogation of the partner and more withdrawal.
Overall, N. C., & Sibley, C. G., Attachment and dependence regulation within daily interactions with romantic partners (2009)(opens in a new tab)What this does not show
Seventy four New Zealand undergraduates, so both the sample and the relationship stage are narrow. Daily co-occurrence establishes that the responses travel together within a day, not which one causes the other, and everything is self-reported by one member of the couple.
- Moderate evidence500 individuals, plus 41 couples across 42 days
Attachment avoidance was associated with more aversive feelings and cognitions during sex, while attachment anxiety was associated with a more ambivalent construal of it, across a large self-report sample and a forty two day couples diary.
Birnbaum, G. E., Reis, H. T., Mikulincer, M., Gillath, O., & Orpaz, A., When sex is more than just sex: Attachment orientations, sexual experience, and relationship quality (2006)(opens in a new tab)What this does not show
Correlational, and the diary component involved forty one couples, which is small for detecting anything subtle. Conducted in Israel with young adults, and self-reported sexual experience is subject to recall and presentation effects.
- Limited evidence47 female undergraduates
College women told that men might like them a lot or only an average amount reported greater attraction than women told the men definitely liked them a lot.
Whitchurch, E. R., Wilson, T. D., & Gilbert, D. T., "He loves me, he loves me not...": Uncertainty can increase romantic attraction (2011)(opens in a new tab)What this does not show
Forty seven participants, all women, all undergraduates, rating profiles of strangers they would not meet, so no real prospect of rejection existed. The design cannot speak to established acquaintance, and the effect has not held up cleanly in later work.
- Moderate evidenceTwo studies of female participants
A later set of studies separating affective from behavioural attraction did not reproduce the original result. Uncertainty raised the wish to meet and talk with someone but not liking or sexual desire, and in a within-participant design certainty of being liked increased attraction.
Montoya, R. M., Faiella, C. M., Lynch, B. P., Thomas, S., & DeLuca, H. K., Further Exploring the Relation Between Uncertainty and Attraction (2015)(opens in a new tab)What this does not show
Also conducted with young women rating profiles rather than in live interaction, so it inherits much of the artificiality of the study it examines. It shows the original effect is fragile and condition-dependent rather than establishing that uncertainty never matters.
- Strong evidenceMultiple experimental studies
Reciprocity of liking is a robust effect: learning that another person likes you increases liking for them, operating through affective, cognitive and behavioural routes, with perceived benevolent intentions doing much of the mediating.
Montoya, R. M., & Insko, C. A., Toward a more complete understanding of the reciprocity of liking effect (2008)(opens in a new tab)What this does not show
Largely laboratory studies of initial attraction among students, which is a poor model of a developing relationship. Establishes a general tendency rather than ruling out individuals for whom it reverses, which is precisely the case in question here.
- Moderate evidenceParticipants across seven speed-dating events
In speed-dating research, unique desire from a specific person prompted unique desire in return, while people who experienced non-selective desire for many partners were desired less in return.
Eastwick, P. W., Finkel, E. J., Mochon, D., & Ariely, D., Selective versus unselective romantic desire: Not all reciprocity is created equal (2007)(opens in a new tab)What this does not show
Brief four-minute encounters among university students, measuring initial desire rather than anything about how relationships develop. The selective and non-selective components are statistically separated in a way that has no clean equivalent in ordinary experience.
- Strong evidence203 assessed daily and 388 assessed weekly
Two intensive longitudinal studies, one assessing attachment daily for thirty days and another weekly for a year, found the pattern of stability best fitted a model with a stable underlying factor beneath genuine temporary variation.
Fraley, R. C., Vicary, A. M., Brumbaugh, C. C., & Roisman, G. I., Patterns of stability in adult attachment: An empirical test of two models of continuity and change (2011)(opens in a new tab)What this does not show
Covers a month and a year respectively, which is short relative to claims about lifelong patterns. Repeated self-report can itself shape responding, and the samples were drawn largely from student and internet populations.
- Strong evidenceOver one thousand participants
Item response theory analysis of adult attachment questionnaires found responses are best represented by continuous dimensions rather than by discrete categories.
Fraley, R. C., Waller, N. G., & Brennan, K. A., An item response theory analysis of self-report measures of adult attachment (2000)(opens in a new tab)What this does not show
Concerns how the questionnaires behave and how respondents distribute across them, not what avoidance predicts. It rules out types as an accurate description of the data and leaves the size of any real-world association entirely open.
How we rate evidence strength
- Strong evidence
- Multiple independent studies, including replications or meta-analyses, point the same way.
- Moderate evidence
- Several studies support this, but samples are limited or findings vary by population.
- Limited evidence
- Early, small, or single-study evidence. Treat this as a reasonable hypothesis, not a settled fact.
- Contested
- Credible researchers disagree, or the evidence points in conflicting directions.
Comparative view
Different perspectives
Traditions and disciplines answer this differently. We describe what each one teaches. We are not telling you which is right, and no tradition speaks with a single voice.
What the Evidence Supports
The asymmetry here is deliberate. The avoidance account draws on a large and coherent literature, though one built substantially on self-report and cross-sectional designs, so the confident statement is that the tendency exists and is measurable, not that it explains any particular case. The uncertainty finding is much weaker: one small experiment, widely amplified, and a follow-up that reproduced only a fragment of it. Popular writing has these two backwards, treating the fragile finding as the headline because it is the more quotable. It is also worth noting what is missing. Nobody has run a study on exactly this question, because there is no clean way to manipulate the moment someone starts to be loved back, so everything above is assembled from adjacent findings and should be read as a plausible account rather than a demonstrated one.
Therapist Perspective
Clinicians who see this presentation commonly report that the interesting material is not the loss of interest but what immediately preceded it, and that clients can often locate a specific moment: a plan made further ahead than usual, a key offered, a sentence that assumed a future. Working on it usually means slowing that moment down rather than debating whether the person was right for them, since the answer to that question tends to change with the state rather than the other way round. Practitioners also describe a common obstacle, which is a client arriving already fluent in the terminology and using it to close the subject rather than open it. This is what practitioners report from consulting rooms, not something established under controlled conditions.
Secular Perspective
Stripped of any moral framing, this is a description of a tendency, and a description is not a permission. You are not required to want anyone, no explanation is owed for ending things, and a feeling that has gone cannot be produced on demand. What does not follow is the further step people take, which is treating the pattern as a fact about themselves that the other person now has to accommodate. The other person is not participating in your pattern; they are just someone who said yes and then found the ground moving. On this view the honest response is to name where you are quickly, absorb the awkwardness of doing so, and stop taking the attention of someone you have already left in every respect except the one that would be uncomfortable to admit.
Practical steps
What you can actually do
Check whether the drop tracks the person or the moment. If it lands at the same point with people who have nothing in common, the pattern is about the point, not the people.
Watch what shows up alongside the flatness. Sudden clarity about a partner's flaws immediately after they commit is a recognisable part of the sequence rather than fresh information.
Separate the feeling from what you do next. You are not obliged to manufacture interest, and you are obliged to say where you are instead of going quiet and letting them work it out.
Notice how much of the earlier intensity was attention rather than wanting. Not knowing keeps someone on your mind, and being on your mind is easy to misread as desire.
Give it longer before concluding. Interest that dips when closeness arrives sometimes returns once the situation stops being novel, and leaving at the first flat week guarantees you never find out.
Worth unlearning
Common misconceptions
That the earlier feeling must have been fake. Deactivation is a response to closeness becoming real, not evidence that the attraction was never there.
That people simply want what they cannot have. The finding behind this slogan came from one small study that later work did not cleanly reproduce, and reciprocated liking usually increases attraction rather than killing it.
That this makes you an avoidant. Attachment is measured as a position on continuous dimensions by self-report questionnaire, not as a category people belong to.
That the pattern is permanent. Intensive longitudinal studies find a stable underlying tendency with genuine movement around it, which is not the same as fixed.
That naming the pattern settles anything with the other person. It accounts for the change in feeling and has nothing to say about how you behave once it changes.
The short version
Key takeaways
Deactivating strategies are triggered by closeness and dependence rather than by anything about the specific person, which explains the timing.
ResearchDaily diary research finds more avoidant people report lower regard, more derogation and more withdrawal on days when they feel more dependent.
ResearchReciprocated liking normally increases attraction, so this pattern departs from the usual finding rather than illustrating a general rule.
ResearchThe claim that uncertainty drives attraction rests on a small study that follow-up work did not cleanly reproduce.
ResearchAttachment avoidance is a continuous self-reported dimension, reasonably stable but not fixed, and is not a clinical category.
By definitionExplaining why interest faded says nothing about how the other person should be treated afterwards.
By definition
Worth saying
When to get professional help
It is worth taking to a therapist if this has repeated across several relationships and is now costing you something you actually want, or if you can see the sequence coming and cannot interrupt it from the inside. That is precisely the kind of pattern individual therapy is suited to, and it does not require you to have identified the correct label first. It is also worth raising if the drop in feeling arrives with contempt or cruelty toward someone who has done nothing except like you, since that part is behaviour rather than emotion and is more changeable than it feels. If you find yourself unable to be alone and unable to stay, that combination tends to be more painful than either alone and is a reasonable thing to bring to a first session. One relationship that cooled is not a pattern and needs no clinical attention at all.
Still wondering
Related questions people ask
Not on the strength of recognising yourself in a description. Avoidance is a position on a continuous scale that everyone occupies somewhere, assessed by questionnaire rather than by matching a profile. Recognising the pattern is useful information about a tendency; it is not an assessment and not a category you have joined.
Community
What other people say
These are readers describing their own relationships, not professionals. Take them as experience, not advice, and please do not name anyone.
No one has shared their experience yet. If you have been here, you would be the first, and probably the most useful thing on this page.
Check our work
Sources
- [1]Attachment Theory and Affect Regulation: The Dynamics, Development, and Cognitive Consequences of Attachment-Related Strategies(opens in a new tab)
Mikulincer, M., Shaver, P. R., & Pereg, D. · Motivation and Emotion · 2003
- [2]Attachment and dependence regulation within daily interactions with romantic partners(opens in a new tab)
Overall, N. C., & Sibley, C. G. · Personal Relationships · 2009
- [3]When sex is more than just sex: Attachment orientations, sexual experience, and relationship quality(opens in a new tab)
Birnbaum, G. E., Reis, H. T., Mikulincer, M., Gillath, O., & Orpaz, A. · Journal of Personality and Social Psychology · 2006
- [4]"He loves me, he loves me not...": Uncertainty can increase romantic attraction(opens in a new tab)
Whitchurch, E. R., Wilson, T. D., & Gilbert, D. T. · Psychological Science · 2011
- [5]Further Exploring the Relation Between Uncertainty and Attraction(opens in a new tab)
Montoya, R. M., Faiella, C. M., Lynch, B. P., Thomas, S., & DeLuca, H. K. · Psychologia · 2015
- [6]Toward a more complete understanding of the reciprocity of liking effect(opens in a new tab)
Montoya, R. M., & Insko, C. A. · European Journal of Social Psychology · 2008
- [7]Selective versus unselective romantic desire: Not all reciprocity is created equal(opens in a new tab)
Eastwick, P. W., Finkel, E. J., Mochon, D., & Ariely, D. · Psychological Science · 2007
- [8]Patterns of stability in adult attachment: An empirical test of two models of continuity and change(opens in a new tab)
Fraley, R. C., Vicary, A. M., Brumbaugh, C. C., & Roisman, G. I. · Journal of Personality and Social Psychology · 2011
- [9]An item response theory analysis of self-report measures of adult attachment(opens in a new tab)
Fraley, R. C., Waller, N. G., & Brennan, K. A. · Journal of Personality and Social Psychology · 2000