Introduction
“Chemistry between people” describes a feeling that the interaction between two people flows particularly well. It can feel like immediate attraction, security, curiosity, familiarity, sexual excitement or a feeling of being on the same wavelength. The expression is used romantically, but also about friendship, cooperation, teaching, leadership, therapy and meetings between health personnel and patients.
Scientifically, this is not one chemical reaction, nor the action of one particular “love molecule.” The experience arises through a dynamic interaction between sensory impressions, the brain’s reward systems, the body’s stress regulation, previous relationship experiences, personality, culture, communication and the specific situation. Therefore, what happens between two people cannot be reduced to the characteristics of each individual. A separate, dyadic component arises in the meeting between these two people. Studies of speed dating show that attraction is influenced by both how generally attractive a person is perceived, how easily a participant is attracted to others, and a unique compatibility between the couple in question.
The purpose of this article is to examine the phenomenon’s historical development, biological and psychological mechanisms, practical significance and future research directions. The target group is the general public, teachers and professionals in health, psychology, management and human interaction.
1. Historical Context: From Philosophy to Relational Science
1.1 Early performances and precursors
Humans have been trying to understand attraction and love long before modern psychology and neuroscience emerged. The philosophers of antiquity discussed, among other things, desire, friendship, affection and community. The explanations were mainly philosophical, ethical or religious, not based on controlled observations.
In the 1800s, the theory of evolution introduced a new perspective. Sexual selection could help explain why individuals prefer certain traits in potential partners. Later psychological directions emphasised unconscious motives, bodily activation, learning and early experiences of care.
These early models were important, but they often mixed observation, culture, and speculation. It was not until the 1900s that researchers began to systematically measure attraction, closeness and love.
1.2 Emergence and key milestones
An important milestone came with studies of physical proximity and contact opportunities. Festinger, Schachter and Back investigated friendship formation in a residential area and showed that architecture and the likelihood of repeated encounters affected who became friends. People who lived close to each other or naturally passed each other frequently were more likely to develop contact. This formed the basis for the proximity or propinity principle.
John Bowlby’s attachment theory from the 1960s and 1970s described how children develop emotional bonds with caregivers, and how such bonds contribute to security and the regulation of stress. The theory was later transferred to adult romantic relationships. In 1987, Hazan and Shaver showed how romantic love could be understood as an adult attachment process, where security, anxiety and avoidance affect how closeness is experienced and handled.
In 1970, Zick Rubin developed scales to distinguish between liking and loving. Thus, romantic love was made into a phenomenon that could be empirically investigated, even though questionnaires naturally only measure parts of a complex experience.
In 1986, Robert Sternberg proposed his triangle model, in which love is described through three components:
Intimacy encompasses closeness, openness, and the feeling of being understood.
Passion encompasses bodily and romantic activation.
Decision and commitment include the desire to maintain the relationship.
The model is a psychological model of organization, not a law of science. It has nevertheless been useful in describing how relationships can contain different combinations of emotional closeness, sexual motivation and long-term commitment.
1.3 Influential academic directions
Several research directions have shaped today’s understanding.
Social psychology investigated how proximity, similarity, reciprocity, attractiveness, status and self-disclosure affect contact. A meta-analysis of self-disclosure found that sharing personal information and being liked are part of a reciprocal system: we share more easily with people we like, we can like people better after they’ve shared something personal, and we can be liked better when we share in an appropriate way.
The attachment research examined how previous relationship experiences affect expectations of closeness, rejection and support. Attachment is not an immutable personal “type”. It can vary between relationships and change through experiences.
Evolutionary psychology investigated possible functions of partner preferences and mating. Such models can explain some statistical patterns, but must be used with caution. Human partner formation is also influenced by culture, economy, social organization, individual learning, and personal values.
Neuroscience and psychoneuroendocrinology began to investigate how the brain’s reward systems, hormones, stress responses, and attachment mechanisms are involved in love and social contact. Imaging studies found that romantic love involves networks related to motivation, attention, and reward learning, but also cortical networks for social understanding and self-representation.
1.4 Key development phases
The history of research can be roughly divided into four phases.
The first phase described love philosophically and culturally. The other measured individual attitudes, characteristics and preferences. The third examined the brain, hormone systems and attachment. The fourth, which characterizes current research, studies the actual interaction between people through speed dating, long-term studies, video coding, physiological measurements, machine learning and simultaneous registration of activity from two brains.
The development thus moves from the question “What qualities make a person attractive?” to a more precise question: “What happens when these two people’s senses, expectations, nervous systems and actions affect each other over time?”
2. Current relevance and scientific understanding
2.1 The current academic landscape
A modern explanation of human chemistry must move between several levels.
At the molecular level , nerve cells release neurotransmitters and neuromodulators. These molecules bind to receptors in cell membranes and alter nerve cell activity, signal processing, and plasticity.
At the network level , brain regions collaborate for reward, memory, attention, social assessment, body perception and emotion regulation.
At the body level , heart rate, breathing, muscle tension, pupil size, skin conduction and hormonal activity change, among other things.
At the psychological level, the signals are interpreted as interest, security, nervousness, attraction or skepticism.
On a relational level , the other person responds to these reactions. This creates feedback loops: A smile triggers a smile, openness is met with openness, or uncertainty creates more uncertainty.
Dopamine and motivated attention
Dopamine, with the molecular formula C₈H₁₁NO₂, is a catecholamine that is involved in motivation, learning and the expectation of reward. It is more accurate to refer to dopamine as a signal for motivation and significance than as a pure “pleasure molecule”.
When a person gains high emotional significance, cues such as their face, name, voice, or a message can become strong attention magnets. Imaging studies of intense romantic love have shown activity in dopaminergic reward and motivation systems. This does not mean that the brain has its own delimited love center. It uses existing systems for motivation and learning to prioritize a specific person.
Norepinephrine, adrenaline and bodily activation
Norepinephrine and epinephrine contribute to alertness and sympathetic activation. The heart can beat faster, attention is sharpened, and digestive activity can be altered. This can be labeled as “butterflies in the stomach”.
However, the same physiological activation can occur in both attraction and agitation. The body does not supply a finished label. The brain interprets the activation based on the situation, expectations and the person’s previous experiences.
Cortisol and stress regulation
Cortisol is a steroid hormone that is synthesized from cholesterol. Because it is fat-soluble, it can cross cell membranes and influence gene expression through glucocorticoid receptors.
Early romantic attraction can be both rewarding and stressful. Uncertainty about reciprocity, strong expectations and increased attention to the relationship can activate the body’s stress system. In safe and stable relationships, on the other hand, the partner can contribute to the regulation of stress. The research therefore describes love as both a possible source of activation and a possible regulatory mechanism.
Oxytocin and vasopressin
Oxytocin is a small peptide consisting of nine amino acids, stabilized by a disulfide bond between two cysteine residues. The molecule is mainly produced in the hypothalamus and is known for its roles in childbirth and breastfeeding, but is also involved in social attention, attachment and stress regulation.
The popular name “cuddle hormone” is misleading. Oxytocin does not automatically make people affectionate or trusting. The effect depends on context, person, expectations, gender, previous experiences and what social signals are present. A meta-analysis of endogenous oxytocin found significant methodological variations and no simple, consistent hormone response to social interaction. Recent models therefore describe oxytocin as a regulator of social significance and behavioral flexibility rather than as a universal love drug.
Vasopressin is a closely related peptide. Both the oxytocin and vasopressin systems can influence attachment, social memory, and protective behavior, but much of the detailed mechanistic knowledge stems from animal models. Transmission to humans must therefore be done with caution.
Endogenous opioids
Endorphins and other endogenous opioids can contribute to social comfort, pain relief, and the calm satisfaction often associated with safe, long-term closeness. Thus, intense early motivation and calm attachment may have partially different biological profiles, even if the systems overlap.
2.2 Dominant patterns in research
Chemistry is both individual and dyadic
Some people are perceived as attractive by many. Others tend to become interested in people in general. But a significant part of the attraction is unique to the individual couple. In an analysis of more than 6600 quick date meetings, first impressions of both overall attractiveness and unique compatibility could predict later messages, dates, and romantic interest.
This explains why one person can be impressive “on paper” without creating chemistry, while another can arouse strong interest without matching a predefined wish list.
Perceived similarity is often more important than objective similarity
Equality in values, humour, lifestyle and form of communication can reduce social insecurity and make cooperation easier. But research distinguishes between actual similarity and perceived similarity. We may perceive a person as similar to us because we already like them, and we may overestimate similarity in good relationships.
In some controlled studies, attitudinal similarity increases attraction, while quick date studies don’t always find that measurable similarity predicts the first spark.
A worldwide study from 43 countries found that ideal partner preferences had some predictive value, but the effect was greatest when the ideals reflected traits most people consider desirable. The unique match between a person’s particular ideals and their partner’s characteristics was weaker.
Reciprocity reinforces interest
It is often attractive to notice that the other person is interested. But interest is strongest when it is perceived as selective. If someone seems equally interested in everyone, the signal may lose value.
At the neurobiological level, reciprocity can reduce uncertainty and at the same time make social contact more rewarding. On a psychological level, it supports self-esteem, security and the expectation of further contact.
Synchronisation and co-regulation
People who communicate well can coordinate their gaze, tone of voice, speech rate, facial expressions, gestures, and pauses. Heart rate, skin conduction and certain patterns in brain activity can also show temporal correlation.
This does not mean that two brains “send thoughts” to each other. Synchronization can occur because both react to the same events, follow each other’s signals, and mutually adjust their behavior. Studies have found correlations between social contact, positive affect, and neural synchronization, and recent meta-analyses examine such patterns in close relationships.
Smell has an effect, but “human pheromones” have not been proven in a robust way
Body odor can contain information influenced by skin bacteria, diet, immune activity, hygiene, hormones, and genetics. Smells can therefore affect closeness and attraction.
However, this is not the same as identifying a specific human sex pheromone that automatically creates desire. Critical reviews have not found robust evidence that the marketed steroids androstadienone, androstenone, androstenol or estratetraenol act as established human pheromones. Human chemical communication is a real field of research, but the mechanisms are more complex than the advertisement for “pheromone perfume” suggests.
2.3 Today’s challenges and controversies
A key challenge is that the word “chemistry” encompasses several different phenomena. Sexual attraction, trust, friendly warmth, intellectual stimulation and secure attachment can occur at the same time, but do not have to do so.
Another challenge is the measurement methods. Questionnaires are influenced by self-awareness and social desirability. Hormone levels in saliva or blood do not necessarily reflect their concentration in specific parts of the brain. Functional MRI detects changes in blood oxygenation, not direct secretion of dopamine or oxytocin.
Many older studies used small and relatively homogeneous samples, often young, heterosexual students from Western countries. It limits generalizability. Recent research uses larger, more culturally diverse samples, but variation in sexual orientation, functional ability, age and form of relationship is still insufficiently represented.
It is also important to distinguish between intensity and quality. Strong activation can be due to secure attraction, but also unpredictability, fear of rejection or conflict. An intense feeling is therefore not automatic documentation of compatibility, safety or future stability.
2.4 Data and societal significance
The WHO’s Commission on Social Connection reported on June 30, 2025, that approximately one in six people globally experience loneliness. The report estimated that loneliness is linked to around 871,000 deaths annually. Such figures describe statistical relationships at the population level and do not mean that loneliness alone can be identified as the cause of death in an individual.
A meta-analysis of 90 prospective cohort studies found that social isolation was associated with about a 32 percent higher risk of death from all causes, while loneliness was associated with about a 14 percent higher risk. Observational studies alone cannot prove that lack of contact causes the entire risk increase, but the results support that social connection is relevant to public health.
At the same time, digital platforms have changed how people meet. A US survey conducted in 2022 and published in 2023 found that 30 percent of adult Americans had used dating sites or dating apps. Among adults in a relationship, about one in ten stated that they had met their current partner through such a service. The figures apply to the United States and cannot simply be transferred to Norway or the rest of the world.
3. Practical applications and consequences
3.1.1 Dating and romantic relationships
The research suggests that dating sites can be good at increasing the number of possible encounters but weaker at predicting the unique chemistry before two people have actually communicated.
In one study, researchers applied machine learning to extensive information about participants’ characteristics and partner preferences. The models were able to predict to some extent who would generally be considered attractive, but were unable to meaningfully predict which two people would feel a unique mutual attraction.
Practically, this implies that dating sites should be considered meeting tools, not chemical compatibility labs. A profile form can reveal clear incompatibilities in values or life goals, but cannot fully replicate voice, timing, humour, smell, eye contact, spontaneity or the feeling of mutual response.
For established couples, “chemistry” is not just something you either have or lack. Positive experiences can change the brain’s expectations of the partner. Attentive listening, appropriate self-disclosure, physical closeness, humor, new joint activities, and constructive conflict management can build stronger connections between the partner and experiences of security and reward.
That doesn’t mean that all relationships should be repaired. In the case of violence, control, threats or persistent violations, safety must weigh more heavily than the desire to restore a sense of chemistry.
3.1.2 Friendship and social health
Friendships are often developed through repeated contact, availability and joint activities. This is a practical explanation of why school, workplace, neighbourhood, volunteering and organised leisure activities are important social arenas.
At the molecular level, repeated safe encounters give the brain more opportunities to update its prediction of the other person. A stranger gradually goes from being insecure to being known. Positive experiences strengthen memories and expectations related to security, support and rewards.
It is therefore often more realistic to facilitate regular contact than to try to “find chemistry” through one perfect meeting. In experiments with live interaction, familiarity has been shown to increase attraction and social interest.
3.1.3 Working life and teams
In working life, “good chemistry” is often used to refer to trust, flow and low social friction. This can improve information sharing and coordination. Meta-analyses find correlations between group cohesion and performance, especially in tasks that require close collaboration.
At the same time, chemistry can become a source of bias. Recruiters and managers may prefer people who are similar to themselves in style, background, and interests. A pleasant conversation is not necessarily the same as a high level of competence.
In recruitment, relational impression should therefore be combined with structured interviews, defined assessment criteria and work samples. In teams, the goal should be psychological safety and functional cooperation, not that everyone should have the same personality.
Good professional chemistry can also occur between different people. Complementary knowledge, respect and clear roles may be more important than spontaneous equality.
3.1.4 Health care and psychotherapy
Chemistry between patient and therapist can be understood as trust, empathy, understanding and agreement on goals and working methods. In psychotherapy, this is often referred to as the therapeutic alliance.
A meta-analysis of 295 independent studies with more than 30,000 patients found a stable association between therapeutic alliance and treatment outcome across multiple therapies. This does not mean that a good relationship alone treats all disorders, but that cooperation can affect openness, motivation, attendance, expectations and implementation of the treatment.
In somatic medicine, clear, empathetic communication is also associated with trust, satisfaction, information understanding and certain health outcomes.
The practical lesson is that relational competence is not a decoration around the “real” treatment. It can be part of the mechanism that ensures that professionally correct treatment is actually understood and followed.
3.2 Overall impact assessment
At the individual level, good interpersonal contact can affect well-being, stress regulation, motivation and belonging.
At the relationship level, it can facilitate communication, mutual support, and conflict resolution.
At the organizational level, it can contribute to collaboration and information flow, but also create favoritism and groupthink.
At the societal level, social connection is relevant for public health, education, working life and the design of physical and digital meeting places.
Chemistry is therefore not just a romantic curiosity. It is an everyday expression of mechanisms that influence how people form and maintain social systems.
3.3 Two immersive cases
Case 1: Speed dating and the unpredictable spark
Before a quick date meeting, researchers can measure personality, intelligence, interests, physical attractiveness, attachment patterns, and partner preferences. This makes it possible to predict some of who will generally be well-liked.
Still, it is far more difficult to predict which particular participant one person will like more than everyone else. The unique response is created during the interaction: how a comment is received, whether the humor hits home, how pauses are handled, and whether the interest is perceived as mutual.
The case shows that chemistry is a process, not just a comparison of two static profiles.
Case 2: The therapeutic alliance
Two therapists can use the same treatment manual, but create different results because they differ in their ability to understand the patient’s perspective, repair misunderstandings, and establish a credible working alliance.
Here, the “chemistry” does not necessarily consist of spontaneous similarity or strong emotional attraction. It can be built through precision, respect, predictability and the experience that the therapist takes the patient’s experiences seriously.
The case illustrates that functional chemistry can be trained and developed, even when it does not occur immediately.
4. Future consequences, trends and research
4.1 Expected research directions
Future research is likely to move away from isolated measurements of individuals and towards continuous measurements of interactions.
Long-term studies can examine how first impressions develop into friendship, love or distance. Instead of just asking if two people have chemistry, researchers can record when it occurs, what events strengthen it, and how it is affected by conflict and repair.
Larger, culturally diverse studies will be able to distinguish between universal biological mechanisms and norms that are shaped by family, culture, gender, economics and technology.
4.2 Technological advances
Hyperscanning involves the simultaneous recording of brain activity in two or more people. EEG and functional near-infrared spectroscopy can be used while people are cooperating, conversing, or solving tasks. Meta-analyses find patterns of interpersonal neural synchronization in close relationships, but the direction of causation is still difficult to determine.
Wearable sensors can detect heart rate variability, conduction, movement, and voice patterns during natural calls. Combined with video data, this can provide a more ecologically realistic picture than short laboratory experiments.
Machine learning can detect composite patterns in large data sets, but previous experiments show that more data does not automatically provide reliable predictions of unique romantic chemistry. Human response is situational, learning and reciprocal. In other words, the model tries to predict a system that changes while it is being observed.
4.3 Opportunities and challenges
The technology can support research on social development, therapy, education, dementia care and communication difficulties. It may be able to identify situations where interaction breaks down or where a treatment alliance is weakening.
At the same time, serious ethical questions arise. Biometric data on gaze, heart rate, voice, emotion, and social response are highly personalized. If employers, dating sites, or insurance companies use such data to classify people as compatible or incompatible, the result may be discrimination and false precision.
An algorithm can also become self-fulfilling. If users are told that they are “97 percent compliant”, the expectation can affect how they interpret the meeting. Conversely, a low algorithmic score can hinder a relationship that could have developed over time and experience.
4.4 The research front
Key questions for further research are:
How do the dopamine, oxytocin, opioid and stress systems interact during natural encounters?
Can physiological synchronisation be distinguished from the fact that two people simply respond to the same external stimulus?
Which forms of chemistry predict secure and stable attachment, and which mainly reflect short-term activation?
How is chemistry affected by age, culture, neurodiversity, previous trauma, sexual orientation and digital forms of communication?
Can social contact be strengthened without biological findings being misused for manipulation or surveillance?
4.5 Long-term perspective
It is unlikely that research will find one universal formula for chemistry between people. The phenomenon arises through several parallel systems that are influenced by both biology and life history.
The models of the future will probably be more precise, but also more humble. They will be able to describe probabilities, mechanisms and risk factors, not decide who a person should “really” love or trust.
Conclusion
Chemistry between people is an emerging phenomenon that occurs when biological, psychological and social processes interact.
Historically, understanding has moved from philosophical and symbolic descriptions to measurable theories of proximity, connection, reward, communication and synchronisation. Research shows that dopamine, oxytocin, vasopressin, stress hormones, and endogenous opioids are involved in social motivation and attachment, but none of the substances act as a simple molecular switch for love.
Today’s relational research emphasizes that chemistry is partly dyadic. It can’t always be predicted from personality tests or wish lists before two people actually meet. Sensory impressions, timing, reciprocity and continuous adaptation create a unique dynamic.
In practical terms, this has implications for dating, friendships, working life, therapy, health care and public health. Good chemistry can support trust and collaboration, but should not be confused with competence, safety or long-term compatibility.
Future research will use more naturalistic measurements, wearable sensors, hyperscanning, and larger cross-cultural studies. At the same time, privacy and the danger of biological reductionism must be taken seriously.
The most precise formulation is therefore that chemistry between people is neither pure magic nor just chemical molecules. It is the subjective experience of a biologically rooted, psychologically interpreted and socially co-created interaction.
Bibliography
Bartz, J. A., Zaki, J., Bolger, N., & Ochsner, K. N. (2011). Social effects of oxytocin in humans: Context and person matter. Trends in Cognitive Sciences, 15(7), 301–309. doi: 10.1016/j.tics.2011.05.002.
Baxter, A., Maxwell, J. A., Bales, K. L., Finkel, E. J., Impett, E. A., & Eastwick, P. W. (2022). Initial impressions of compatibility and mate value predict later dating and romantic interest. Proceedings of the National Academy of Sciences, 119(45), e2206925119. doi: 10.1073/pnas.2206925119.
Beal, D. J., Cohen, R. R., Burke, M. J., & McLendon, C. L. (2003). Cohesion and performance in groups: A meta-analytic clarification of construct relations. Journal of Applied Psychology, 88(6), 989–1004. doi: 10.1037/0021-9010.88.6.989.
Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. Basic Books.
Burenkova, O. V., Dolgorukova, T. A., An, I., Kustova, T. A., Podturkin, A. A., Shurdova, E. M., Talantseva, O. I., Zhukova, M. A., & Grigorenko, E. L. (2023). Endogenous oxytocin and human social interactions: A systematic review and meta-analysis. Psychological Bulletin, 149(9–10), 549–579. doi: 10.1037/bul0000402.
Eastwick, P. W., Sparks, J., Finkel, E. J., Meza, E. M., Adamkovič, M., et al. (2025). A worldwide test of the predictive validity of ideal partner preference matching. Journal of Personality and Social Psychology, 128(1), 123–146. doi: 10.1037/pspp0000524.
Feldman, R. (2017). The neurobiology of human attachments. Trends in Cognitive Sciences, 21(2), 80–99. doi: 10.1016/j.tics.2016.11.007.
Festinger, L., Schachter, S., & Back, K. W. (1950). Social pressures in informal groups: A study of human factors in housing. Harper.
Fisher, H., Aron, A., & Brown, L. L. (2005). Romantic love: An fMRI study of a neural mechanism for mate choice. Journal of Comparative Neurology, 493(1), 58–62. doi: 10.1002/cne.20772.
Flückiger, C., Del Re, A. C., Wampold, B. E., & Horvath, A. O. (2018). The alliance in adult psychotherapy: A meta-analytic synthesis. Psychotherapy, 55(4), 316–340. doi: 10.1037/pst0000172.
Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52(3), 511–524. doi: 10.1037/0022-3514.52.3.511.
Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relationships and mortality risk: A meta-analytic review. PLOS Medicine, 7(7), e1000316. doi: 10.1371/journal.pmed.1000316.
Joel, S., Eastwick, P. W., & Finkel, E. J. (2017). Is romantic desire predictable? Machine learning applied to initial romantic attraction. Psychological Science, 28(10), 1478–1489. doi: 10.1177/0956797617714580.
Luo, S., & Zhang, G. (2009). What leads to romantic attraction: Similarity, reciprocity, security, or beauty? Evidence from a speed-dating study. Journal of Personality, 77(4), 933–964. doi: 10.1111/j.1467-6494.2009.00570.x.
Rubin, Z. (1970). Measurement of romantic love. Journal of Personality and Social Psychology, 16(2), 265–273. doi: 10.1037/h0029841.
Sternberg, R. J. (1986). A triangular theory of love. Psychological Review, 93(2), 119–135. doi: 10.1037/0033-295X.93.2.119.
World Health Organization. (2025). From loneliness to social connection: Charting a path to healthier societies. Report of the WHO Commission on Social Connection.
Wyatt, T. D. (2015). The search for human pheromones: The lost decades and the necessity of returning to first principles. Proceedings of the Royal Society B, 282(1804), 20142994. doi: 10.1098/rspb.2014.2994.
Zhao, Q., Zhao, W., Lu, C., Du, H., & Chi, P. (2024). Interpersonal neural synchronization during social interactions in close relationships: A systematic review and meta-analysis of fNIRS hyperscanning studies. Neuroscience & Biobehavioral Reviews, 158, 105565. doi: 10.1016/j.neubiorev.2024.105565.