Love is not one neurochemical state. It is at least three distinct states, each governed by a different molecule, each producing a qualitatively different subjective experience, and each with a predictable lifespan. The feeling that most people call “being in love” — the obsessive thinking, the elevated energy, the sense that this person completes something fundamental — is the dopamine-dominant phase. It lasts 12 to 18 months. When it ends, most people conclude the love has ended. What has actually ended is one neurochemical phase of a multi-phase process, and the phase that follows — if the couple builds the conditions for it — produces a form of connection that the dopamine phase cannot.
What follows maps the three neurochemical phases, the specific point where most relationships fail during the transition between them, and the three clinical presentations of bonding failure I observe consistently in practice. Each phase requires different behavioral conditions to sustain — and building those conditions deliberately is the difference between relationships that deepen and relationships that dissolve.
The neurochemistry of love is not a metaphor. It is a sequence of measurable states, each with a distinct architecture and a finite duration. The couples who last are not the ones who maintain the first state forever. They are the ones who know how to build the conditions for each phase as it arrives.
I have spent 26 years watching couples arrive at the same bewildering moment. They describe it in remarkably similar language: “Something changed.” “It doesn’t feel the same.” “We lost what we had.” What they are describing, in every case, is a neurochemical phase transition that they were never told was coming. The attraction that felt automatic has become effortful. The person who seemed to read their mind now seems to inhabit a different reality. They interpret this as evidence of a failing relationship. In most cases, it is evidence of a normally functioning brain completing one neurochemical program and requiring the conditions for the next one to activate. The relationship has not failed. The couple has simply reached the end of the phase they were equipped for and the beginning of a phase they were never taught to navigate.
What Neurochemicals Drive the Early Phase of Romantic Love?
The initial phase of romantic love — what psychologist Dorothy Tennov named limerence — is dominated by dopamine and norepinephrine, producing a neurological state that resembles a mild form of obsessive-compulsive disorder more than it resembles contentment. Helen Fisher’s neuroimaging research at Rutgers documented that people in early-stage romantic love show activation of the caudate nucleus and ventral tegmental area — reward and motivation centers — not the brain regions associated with calm attachment or emotional security (Fisher et al., 2005). Early love is a pursuit state. The brain is not registering safety. It is registering something it wants and does not yet have.
Dopamine during limerence follows the reward prediction error model established by Wolfram Schultz: the signal fires most intensely when outcomes are uncertain. In early romance, every interaction carries uncertainty. Will they respond? Do they feel the same way? What will happen next? Each positive response generates a prediction error spike that produces the subjective experience of falling — the swooping, electric quality that people describe as “chemistry.” This is not a measure of compatibility. It is a measure of uncertainty. The brain is doing exactly what it does with any unpredictable reward source: generating massive motivational energy to pursue it.
Simultaneously, serotonin levels decrease during limerence, producing the obsessive thinking that characterizes new love — a neurochemical shift that, when sustained into the bonding phase, can contribute to the patterns that turn marriages sexless. Donatella Marazziti’s research at the University of Pisa found that serotonin transporter levels in people experiencing early romantic love were indistinguishable from those in patients with obsessive-compulsive disorder (Marazziti et al., 1999). The person you cannot stop thinking about is not on your mind because the connection is uniquely profound. They are on your mind because your serotonin system is operating in a state that makes intrusive, repetitive thoughts about a specific stimulus neurologically inevitable.
In my practice, I consistently observe that the couples most vulnerable to subsequent failure are those whose limerence phase was most intense. The neurochemical high was so pronounced that its normalization — which is biologically inevitable — registers as catastrophic loss rather than natural transition. They spent the dopamine phase believing they had found something rare, when what they had found was a normally functioning reward prediction system operating at high amplitude. The higher the amplitude, the more devastating the perceived loss when it normalizes.
What Happens Neurochemically When the “Honeymoon Phase” Ends?
Between 12 and 18 months into a relationship — the range extends to 36 months in cases where external circumstances sustain high prediction error — the dopamine-dominant phase gives way to what should be an oxytocin-dominant phase. This is not a gradual dimming. It is a phase transition: the neurochemical architecture of the relationship reorganizes around a fundamentally different molecule with fundamentally different behavioral requirements.
Oxytocin, produced in the hypothalamus and released through physical closeness, sustained touch, and sexual connection, builds the infrastructure of secure attachment. Kerstin Uvnas Moberg’s research documented that oxytocin activates the calm-and-connection system — a parasympathetic state that reduces cortisol, diminishes threat sensitivity, and produces the felt sense of safety in another person’s presence that is the foundation of long-term pair bonding. This is not the excitement of limerence. It is something neurologically distinct: a quiet, warm, physiologically stabilizing state that deepens with consistent activation over time.
The transition fails when the behavioral conditions for oxytocin bonding have not been built during the dopamine phase. The couple spent the previous year in pursuit mode — intense, novel, occasionally volatile — but did not systematically create the conditions that oxytocin bonding requires: sustained physical closeness without sexual agenda, emotional attunement under stress, reliable presence during vulnerability, the repeated experience of being genuinely known and remaining chosen. When the dopamine signal quiets, there is nothing structurally beneath it. The couple free-falls.
What I see consistently in practice is a specific presentation: one or both partners feel inexplicably disconnected from someone who used to feel essential. Small frictions that limerence smoothed over become prominent. The relationship feels like effort where it previously felt effortless. Both partners often conclude independently that they chose the wrong person. They are almost certainly wrong. They chose someone appropriate. They failed to build the neurochemical architecture that the next phase of love runs on.
What Does Bonding Failure Look Like in Practice?
The failure of oxytocin bonding to replace dopamine excitement clusters into three presentations I have observed across hundreds of couples in 26 years.
The first is the intensity-dependent couple. These partners built their early relationship on high-stimulus experience — frequent travel, intense conflict followed by intense reconciliation, social environments that sustained dopamine through external novelty. When ordinary life reasserts itself, the dopamine quiets, and neither partner has developed the habits of the quieter attachment system. They mistake the absence of intensity for the absence of love. These couples often cycle through breakups and reconciliations, because the threat of loss temporarily reactivates dopamine prediction error, producing the intensity they equate with connection. The reconciliation feels like proof the love is real. It is proof that prediction error has been reactivated. These are not the same thing.
The second is the avoidantly attached partner. This individual’s early neurological history encoded proximity as unreliable or threatening. For them, oxytocin bonding requires tolerating sustained closeness, which activates a stress response that counteracts the bonding signal. The nervous system treats intimacy beyond a certain depth as a threat. The dopamine phase, with its natural cycle of approach and withdrawal, can feel more manageable than the sustained vulnerability that oxytocin bonding demands. When limerence ends and the relationship requires deepening into attachment, this partner pulls back — often without understanding why, often with explanations that locate the problem in the relationship rather than in their own attachment architecture.
The third is serotonin asymmetry — what happens when one partner’s neurochemistry normalizes faster than the other’s. The partner who has moved through the dopamine cliff experiences the other’s continued intensity as neediness or instability. The partner still in dopamine-dominant mode experiences the other’s normalization as withdrawal or lost interest. What is actually occurring is a phase gap: two people in different neurochemical states simultaneously, interpreting each other’s behavior through incompatible neurological frameworks. Without a shared language for this asynchrony, each partner’s adaptive response amplifies the other’s distress. He pulls back because her intensity feels pressuring. She pursues harder because his distance feels like abandonment. The cycle escalates until one or both conclude the relationship is unsustainable.
In my practice, identifying which pattern is operating determines the intervention. Each requires a different approach. But the foundational work is identical: building the behavioral conditions — the consistent physical closeness, the attunement under stress, the reliable emotional presence — that allow the oxytocin system to activate and stabilize. This work is not romantic in the way the dopamine phase was romantic. It is precise, deliberate, and effective.
How Do You Build Love That Survives the Phase Transition?
The couples who build durable relationships are not the ones who maintain limerence — that is neurologically impossible. They are the ones who successfully navigate the transition from dopamine to oxytocin, from the excitement of not-yet-knowing to the security of being genuinely known. This transition does not happen by accident. It happens by creating the specific conditions each neurochemical system requires.
During the dopamine phase, while the motivational energy is high and the partners are highly attuned to each other, the critical investment is building the behavioral patterns that oxytocin bonding will require: consistent physical touch that is not always sexual, emotional disclosure under stress rather than only during calm, reliability of presence during the partner’s difficult moments, and the willingness to be seen accurately rather than idealized. These behaviors feel less exciting than the dopamine-driven intensity. They are more important. They are the foundation the relationship will stand on when the dopamine phase ends.
During the oxytocin transition, the critical task is tolerance — specifically, tolerating the neurochemical shift without interpreting it as loss. The dopamine will diminish. The obsessive thinking will quiet. The partner will become predictable. None of this is failure. All of it is prerequisite for the deeper, more stable state that oxytocin builds. The couples who interpret this transition correctly — as the relationship graduating from pursuit to partnership — navigate it with relative ease. The couples who interpret it as evidence that something has gone wrong begin the withdrawal patterns that lead to the presentations described above.
Long-term attachment introduces additional neurochemistry: vasopressin, which promotes partner-specific bonding and protective behavior — the same neurochemical architecture that determines how dopamine shapes relationship patterns — and endogenous opioids, which produce the warm, comfortable, physiologically stabilizing experience of being with someone who has become neurologically integrated into your stress regulation system. This phase — when functioning well — is the most rewarding. Not because it produces the highest neurochemical amplitude, but because it produces the most stable, durable, and physiologically beneficial form of human connection available.
Couples in this phase describe a form of connection that is quiet but structurally different from anything the dopamine phase produced — a sense of being regulated by the other person’s presence, of the nervous system settling into a baseline that is measurably calmer when the partner is nearby. In my practice, I observe that partners who have reached this phase show reduced cortisol reactivity during conflict and faster autonomic recovery after stress. The relationship has become, at the neural level, part of each person’s stress regulation infrastructure — not a source of excitement, but a source of physiological stability that no other relationship in their life replicates.
For a comprehensive framework on how the dopamine reward system operates across all phases of romantic attachment — including the prediction error mechanisms that drive both the intensity of early love and the extinction of desire in long-term partnerships — the full science is covered in The Dopamine Code (Simon & Schuster, June 2026).
What a Neuroscientist Does Differently
When a couple comes to me reporting that they have “lost the feeling,” I do not attempt to recreate the feeling they lost. That feeling was dopamine-driven limerence, and it has a biological expiration date. Instead, I assess which phase transition is stalled, which neurochemical system needs activation, and what specific behavioral conditions are missing.
Using EQ Architecture Protocol™, I work with couples during the live moments when the phase transition is failing — when the avoidant partner’s proximity tolerance hits its limit and the withdrawal impulse fires, when the intensity-dependent couple encounters the flatness of ordinary Tuesday and the panic of perceived loss arrives, when the partner still in dopamine mode reaches for reassurance and the partner in oxytocin mode cannot provide the intensity being sought. These are the moments when the relational architecture is most plastic and most responsive to restructuring. The intervention is not retrospective. It is neurologically present-tense.
Relationships & Dating — MindLAB Locations
References
Fisher, H. E., Aron, A., & Brown, L. L. (2005). Romantic love: A mammalian brain system for mate choice. Philosophical Transactions of the Royal Society B, 360(1463), 2173-2186. https://doi.org/10.1098/rstb.2005.1736
Marazziti, D., Akiskal, H. S., Rossi, A., & Cassano, G. B. (1999). Alteration of the platelet serotonin transporter in romantic love. Psychological Medicine, 29(3), 741-745. https://doi.org/10.1017/S0033291798007946
Uvnas-Moberg, K., Handlin, L., & Petersson, M. (2015). Self-soothing behaviors with particular reference to oxytocin release induced by non-noxious sensory stimulation. Frontiers in Psychology, 5, 1529. https://doi.org/10.3389/fpsyg.2014.01529
Frequently Asked Questions
Is the neurochemistry of love the same for everyone?
The neurochemicals involved are identical across individuals: dopamine, norepinephrine, serotonin, oxytocin, vasopressin, and endogenous opioids all participate in the same sequence. What varies is the amplitude, duration, and phase-transition timing. Individuals with anxious attachment histories tend to produce more intense dopamine responses during limerence and experience more distressing withdrawal when the phase ends. Individuals with avoidant attachment show blunted oxytocin responses that make the bonding phase more difficult to establish. These differences are not personality — they are neurological architecture shaped by early relational experience, and they are modifiable through targeted intervention.
Can you fall in love with the same person twice?
You cannot replicate the full dopamine-dominant limerence phase with someone your brain already has a complete prediction model for, because prediction error — the mechanism driving limerence — requires genuine uncertainty. However, significant life changes, extended separation, or major personal transformation in one partner can partially reset the prediction model, producing attenuated but real dopamine-driven attraction. More importantly, the oxytocin and vasopressin systems that sustain long-term attachment can deepen indefinitely with consistent activation, meaning that the experience of love in year 20 can be more neurologically integrated — and more physiologically stabilizing — than the experience in year one, even without the dopamine intensity.
Why do some couples stay intensely attracted for decades while others lose interest quickly?
Couples who maintain attraction over decades are not sustaining the limerence phase — that is biologically impossible. They are doing two things: successfully transitioning through each neurochemical phase as it arrives, and consistently introducing genuine novelty that periodically reactivates dopamine prediction error within the security of established oxytocin bonding. The combination — deep security with periodic surprise — produces a relational experience that is neurologically richer than either limerence or stable attachment alone. Couples who lose interest quickly have typically failed the dopamine-to-oxytocin transition and have no framework for understanding why the intensity faded.
How do attachment styles affect the neurochemistry of love?
Attachment style directly shapes neurochemical response patterns. Securely attached individuals produce moderate dopamine during limerence and transition smoothly to oxytocin bonding because their early relational history encoded closeness as safe. Anxiously attached individuals produce exaggerated dopamine responses (the limerence feels overwhelming) and struggle with the transition because the dopamine phase resembles the intermittent reinforcement of their early attachment environment. Avoidantly attached individuals may produce normal dopamine during limerence but show blunted oxytocin receptor sensitivity, making the bonding phase feel threatening rather than rewarding. These patterns are not fixed. They are neural architectures that can be restructured through targeted intervention during the moments when the attachment system is active.
Is there a way to speed up the transition from limerence to secure attachment?
The phase transition cannot be accelerated neurochemically — the dopamine system normalizes on its own biological timeline. What can be accelerated is the construction of the behavioral conditions that oxytocin bonding requires. Couples who invest early in consistent physical affection, emotional vulnerability under stress, and reliable presence during difficulty build the oxytocin infrastructure while dopamine is still high. This means that when the dopamine phase naturally concludes, the oxytocin system is already active and the transition feels like a deepening rather than a loss. Couples who spend the entire dopamine phase pursuing intensity without building attachment infrastructure experience the transition as a freefall.
If You Recognize These Patterns
If you recognize the patterns described here — the inexplicable flatness after an intense beginning, the sense of distance that arrived without a clear cause, the conclusion that the love must have been an illusion because it feels so different now — and conventional approaches have not resolved the disconnection, a strategy call identifies which specific phase transition is stalled and what the neurological intervention looks like for your particular architecture.
This article is part of our Intimacy & Bonding collection. Explore the full series for deeper insights into intimacy & bonding.
