Brain Reward System: Circuits, Reinforcement, and Motivation
Author: Ukrainian Psychological Hub · Published: October 12, 2026 · Editorial Policy
The brain reward system is a distributed set of neural circuits that helps organisms learn which outcomes matter, anticipate opportunities, allocate effort, and adapt choices through experience. Its most studied components connect midbrain dopamine neurons with the striatum, prefrontal cortex, pallidum, thalamus, amygdala, hippocampus, and related structures. Reward processing encompasses prediction, reinforcement, approach, valuation, pleasure, and learning, each of which can vary independently. Haber and Knutson's anatomical synthesis maps the core cortico–basal ganglia architecture.
A meal after hunger, praise after difficult work, relief from discomfort, a successful prediction, and a pleasant social exchange can all recruit elements of these circuits. Yet none activates a universal, isolated 'pleasure button.' A reward may generate pleasure, strengthen a behavior, capture attention, or change expectations, with the balance depending on context, physiological state, prior learning, and the task being studied.
What is the brain reward system?
In behavioral neuroscience, reward describes an outcome with the capacity to influence behavior, valuation, or learning under specified conditions. Reinforcement describes a change in the future probability of a response produced by its consequences. Positive reinforcement introduces a consequence that increases responding; negative reinforcement increases responding through removal or avoidance of something aversive. Both are technical learning concepts, and neither is equivalent to 'feeling good.' An event can be intensely pleasant without producing reliable reinforcement in a given setting, and behavior can be reinforced without strong conscious pleasure.
The expression 'brain reward system' is useful shorthand for an interacting collection of circuits. It does not designate a single anatomical organ. The term 'mesolimbic dopamine pathway' usually refers to dopamine projections from the ventral tegmental area (VTA) toward limbic targets including ventral striatum; 'mesocortical' projections extend toward cortical territories. Corticostriatal and pallidothalamic connections integrate these signals with action selection, goals, context, and evaluation. Anatomical boundaries overlap and their functions are neither perfectly modular nor confined to dopamine.
How reward neuroscience developed
The experimental roots of reward-circuit science include the 1950s discovery that animals would repeatedly perform actions producing electrical stimulation at certain brain sites. Researchers subsequently mapped pharmacological and physiological mechanisms, identified dopamine-rich projections, and examined how rewards and reward-predicting cues influence neural activity. These foundational animal experiments established causal intervention techniques, but electrical self-stimulation cannot be assumed to measure the same experience as human happiness.
During the 1990s, recordings from primate midbrain neurons showed that some dopamine responses reflect discrepancies between predicted and delivered rewards. Schultz, Dayan, and Montague (1997) connected those findings to computational learning theory. Later human functional MRI work related reward tasks to activity in ventral striatum and prefrontal regions. Anatomical studies demonstrated that ventral and dorsal corticostriatal systems interact rather than functioning as sealed reward and movement compartments. For a detailed account of the learning computation and its evidence, see Reward Prediction Error: How the Brain Learns From Outcomes.
Subsequent experiments separated incentive motivation from hedonic experience, studied the roles of endogenous opioids and endocannabinoids, and examined the influence of effort, delay, uncertainty, and environmental cues. Contemporary reviews emphasize the diversity of dopamine signals and the possibility that some signals teach sensory and associative features beyond simple scalar value. See Gershman and colleagues (2024) and Kahnt and Schoenbaum (2025).
The major reward circuits and what their components contribute
Ventral tegmental area and midbrain dopamine neurons
The VTA is a heterogeneous midbrain region containing dopamine-producing cells and other neuronal populations. Its dopamine projections reach ventral striatum and several cortical and limbic targets. Individual cell populations differ in inputs, projection destinations, firing patterns, and responses to rewarding, aversive, novel, or behaviorally important events. Dopamine neurons in the substantia nigra also contribute to motivational and action-related signaling. A textbook diagram showing one VTA arrow ending in the nucleus accumbens represents only part of the circuitry.
Some dopamine cells exhibit brief, phasic changes in firing after unexpected events or informative cues; other aspects of dopamine signaling unfold over longer timescales. These dynamics can influence learning, the assignment of incentive significance, and the willingness to mobilize behavior. The observation of a phasic response alone cannot establish that a person enjoyed an event. Schultz's 2016 review explains the temporal complexity of reward-related dopamine activity.
Nucleus accumbens and the ventral striatum
The nucleus accumbens is a major part of the ventral striatum, commonly divided into core and shell territories. It integrates cortical glutamatergic inputs, signals carrying information about context and emotion, and neuromodulatory influences including dopamine. Its activity is relevant to cue-triggered approach, action vigor, reinforcement, behavioral flexibility, and certain forms of reward learning. Different accumbens subregions and cell types may support distinct or even opposing behavioral effects.
The accumbens also figures prominently in pleasure research, but the distinction is crucial: experimental work in rodents localized enhancements of sweet-taste 'liking' to limited hedonic hotspots within its shell, depending on opioid and endocannabinoid signaling. This result concerns measured affective taste reactions in animals and supports a mechanistic distinction between hedonic impact and pursuit; it is not a direct map of every human pleasant feeling. See Mahler, Smith, and Berridge (2007) and Berridge and Kringelbach (2015).
Ventral pallidum, thalamus, and connected basal ganglia loops
The ventral pallidum receives major input from ventral striatum and participates in coordinating motivational and behavioral output. Certain experimentally identified sites have been implicated in hedonic reactions and reward-related motivation, especially in animal studies. The pallidum and thalamus also participate in recurrent loops linking cortex and basal ganglia. These loops integrate information about value, action possibilities, and context, allowing changing goals to influence what an organism does next.
The dorsal striatum, including caudate and putamen, is often examined in relation to action selection, learned action sequences, and habits. Ventral-to-dorsal interactions are relevant to how reward-guided behavior can become more automatic with practice. Such change depends on many learning and control systems; it is an oversimplification to claim that dopamine alone turns voluntary choices into habits.
Orbitofrontal, ventromedial, and anterior cingulate cortices
Orbitofrontal and ventromedial prefrontal regions participate in representing and updating expected outcomes, integrating dimensions of value, and using learned information to guide decisions. The anterior cingulate and related medial frontal circuits contribute to monitoring costs, effort, conflict, and control demands. These are overlapping functional descriptions, not exclusive labels attached to single parcels. Their observed activity depends on task design, timing, sensory properties, learning history, and the available alternatives.
A person choosing between an immediately available small benefit and a larger delayed benefit engages more than a dopamine response: the decision depends on predicted outcomes, timing, uncertainty, physical and cognitive effort, and higher-level priorities. Networks connecting prefrontal regions, striatum, and thalamus are well suited to integrate these signals. Haber and Knutson (2010) describe why anatomy and human imaging must be interpreted together.
Amygdala, hippocampus, lateral habenula, and brainstem
Amygdala circuits help encode the significance and identity of stimuli, particularly when cues become associated with positive or aversive consequences. The hippocampus contributes episodic, relational, and spatial context; the same cue can acquire different value depending on where, when, and with whom it appears. Lateral habenula and interconnected midbrain systems are implicated in signaling omitted rewards or negative outcome information, though these roles cannot be reduced to a single 'disappointment center.'
Reward circuits also communicate with hypothalamic and brainstem systems involved in arousal, physiology, stress, and bodily needs. Hunger, satiety, pain, sleep, threat, and hormones influence reward valuation. This explains why a given outcome changes its attractiveness across situations without requiring a permanent alteration in one dopamine 'level.'
How the reward system works: a sequence of interacting computations
1. Prediction and anticipation
Experience builds predictions about the outcomes of particular cues and actions. A restaurant sign may evoke an expectation of food; a notification sound may evoke uncertainty about an incoming message; a workplace signal may predict praise or criticism. The reward network integrates sensory and contextual information with remembered outcomes. Anticipatory neural activity often reflects an expected opportunity or the importance of an upcoming decision rather than pleasure already being experienced.
2. Valuation and choice
The brain weighs the likely benefits of potential outcomes against their costs, delays, risks, and alternatives. Researchers distinguish outcome value from policy or action value, and learned or cached values from values inferred using a model of the environment. These computational quantities are simplifications of behavior, and distinct models can sometimes fit the same observations. Decision signals emerge across networks, rather than from a single internal currency meter with a directly readable psychological meaning.
3. Action initiation and effort
Obtaining rewards often requires work. Dopamine-sensitive circuits are involved in how vigorously actions are performed and how willing an organism is to pursue a more effortful option. In experimental tasks, altering dopamine systems can shift allocation toward lower-effort alternatives without necessarily removing the ability to appreciate the reward itself. Salamone and Correa (2024) review the activational and effort-related dimensions of motivation.
4. Outcome experience
When an outcome arrives, sensory, interoceptive, emotional, and contextual processes contribute to its experienced quality. Taste, relief, curiosity, social belonging, and achievement are not interchangeable sensations. Hedonic processing includes systems that extend beyond dopaminergic projections, and subjective pleasure cannot be reconstructed from one measurement of neural activation. Pleasure and motivational pursuit are coupled in many situations, but they can diverge.
5. Learning and updating
The difference between expected and actual outcomes can help update future predictions and behavior. The outcome may be better than expected, approximately as expected, or worse than expected. Such differences are often modeled as reward prediction errors and studied in relation to transient dopamine signals. They operate within a richer set of learning processes that include attention, sensory identity, causal structure, exploration, uncertainty, and context.
Reward, reinforcement, motivation, wanting, and liking
Several concepts are often conflated in everyday explanations. 'Liking' refers to hedonic impact; 'wanting,' especially in incentive-salience theory, concerns the motivational pull of a target or cue; and learning concerns acquired associations or updated expectations. These components frequently co-occur in ordinary life, but experiments show that their underlying mechanisms can be partially dissociated. Berridge, Robinson, and Aldridge (2009) offer the influential three-component distinction.
Consider a person who anticipates a food they once found delicious. An associated smell may acquire strong motivational power, increasing the desire to seek the food. After eating, pleasure may be modest because of satiety or changed preference. The learned cue, the drive to pursue, and the consummatory experience are related processes with different determinants. In clinical settings, a similar distinction helps researchers discuss craving, habitual behavior, and loss of pleasure without assuming that all three represent the same deficit.
Incentive-sensitization theory proposes that reward-associated cues can acquire especially strong motivational significance, sometimes disproportionate to hedonic enjoyment. The theory has been influential in addiction research and remains one account among several, including learning/habit, opponent-process, and impaired-control frameworks. Robinson and Berridge (2025) examine the theory and its critiques thirty years after its original formulation.
Dopamine: what it does and what it does not explain
Dopamine is a neuromodulator that influences neural processing across multiple systems. It contributes to reward-related learning, incentive motivation, effort allocation, action selection, and movement, among other functions. Calling dopamine a 'pleasure chemical' obscures these roles and falsely equates release with a conscious positive emotion. Different brain regions, receptor types, time courses, and experimental conditions produce different outcomes.
A useful distinction separates brief phasic changes associated with events from slower changes in local dopamine availability. Researchers also distinguish spikes recorded in neurons from chemical measurements at their terminals. Firing rates, release, receptor activation, and population-level fMRI BOLD responses are related through physiology but cannot be used as interchangeable measurements. Individual neurons and projections are heterogeneous, while dopamine can signal motivationally relevant information even when the event is aversive or unexpected rather than pleasant.
A 2025 systematic review and meta-analysis of 102 randomized placebo-controlled studies in healthy adults found that dopamine-upregulating manipulations were associated with modest improvements in composite reward measures and particular effects on reward learning or sensitivity, reward response vigor, and reward discounting. Serotonin manipulations showed different associations with punishment-related learning and discounting. These group-level pharmacological findings depend on heterogeneous drugs and tasks; they do not yield a personal dopamine score or a clinical test of motivation.
Dopamine is not the only chemical participant. Glutamate transmits substantial excitatory information within corticostriatal circuits; GABA participates in local and long-range inhibitory signaling. Endogenous opioids and endocannabinoids have important roles in some hedonic and consummatory mechanisms, as well as wider brain functions. Serotonin, norepinephrine, acetylcholine, and multiple peptide systems interact with learning, arousal, stress, state, and behavior. The meaning of a signal depends on its receptor, location, timing, and circuit context.
Reward prediction errors and reinforcement learning
In a simple computational model, prediction error equals obtained outcome minus expected outcome. A better-than-expected reward gives a positive error; a fully predicted reward gives a smaller error; an omitted expected reward gives a negative error. In certain laboratory settings, dopamine activity initially increases at an unpredicted reward and later shifts toward a predictive cue as learning develops. This widely replicated pattern motivated temporal-difference accounts of reinforcement learning. Schultz (2016) reviews the basic physiology.
The equation is a model, not a literal measurement of every person's thoughts. Rewards differ in probability, magnitude, timing, sensory identity, and relevance to current needs. A single event can change future action policies and predictions about hidden environmental states. Some dopamine responses reflect information about what reward was expected, not merely how valuable it was. Kahnt and Schoenbaum's 2025 analysis reviews evidence from human and animal research for these richer representations.
Researchers distinguish model-free learning, which updates action values from past consequences, from model-based learning, which uses knowledge of transitions, outcome identities, and causal structure to evaluate possible futures. Real behavior combines strategies, and experimental tasks rarely isolate them perfectly. Reward prediction error is important enough to warrant a separate canonical treatment of its mathematics and neural models; this article establishes how it functions within the wider reward circuit.
Pavlovian, instrumental, and habit-related reward learning
Pavlovian learning changes how predictive cues influence responses. Instrumental learning changes the likelihood of actions depending on their consequences. A cue predicting something desirable may elicit approach even when approaching is counterproductive, while an action can be repeated because previous consequences made it advantageous. The two processes interact through mechanisms studied as Pavlovian-to-instrumental transfer, but they are experimentally distinguishable.
With repeated experience, behavior may rely increasingly on automatic stimulus–response control. Yet habit formation is affected by context, training, stress, reward devaluation, and executive control; repeated behavior is not automatically a habit in the scientific sense. The related English Hub article How Habits Form: Cues, Repetition, and Automaticity develops the psychological evidence. Here the focus is on how corticostriatal systems interact during action learning.
Reward-learning studies also distinguish reinforcement from explicit incentives. Offering money or praise can affect current performance, interest, expectations, and social meaning in different ways. The psychological question of how external incentives influence intrinsic motivation has its own research tradition, reviewed in How Rewards Affect Motivation. Neural mechanisms help explain facets of these effects without replacing motivational or developmental psychology.
Why reward signals change with hunger, uncertainty, effort, and social context
A reward's value is state-dependent. Food generally changes value after satiety; a delayed benefit may lose influence when the person is exhausted or pressed for time. Uncertainty can increase attention and exploration but also produce avoidance. Reward tasks therefore often manipulate probability, magnitude, delay, cost, effort, or learning contingency independently, because treating them as one psychological dimension would erase important differences.
Social rewards—including affiliation, recognition, and cooperation—can recruit some overlapping valuation circuitry with money, food, or other outcomes. Shared population-level activation does not prove these experiences are identical or that every social interaction is processed by one universal 'social dopamine circuit.' Interpretations must account for meaning, culture, relationships, and the specific task design. Likewise, a pleasurable music passage or successful problem-solving moment may involve combinations of valuation, emotion, memory, prediction, and bodily responses.
Laboratory studies sometimes show neural correlations between surprise or recent reward history and momentary happiness ratings. For instance, Rutledge and colleagues (2014) modeled self-reported momentary well-being in reward tasks and linked task variables to striatal BOLD responses. Such research provides a formal account of one experimental phenomenon; it does not establish a general formula for happiness, meaning, or long-term life satisfaction.
Reward circuitry and addiction: a mechanistic connection
Substance use can interact with circuits involved in reinforcement, learned cues, stress, and behavioral control. Different substances have different direct pharmacological mechanisms, and repeated exposure can produce neuroadaptations. Addiction is a clinical and behavioral syndrome involving multiple interacting mechanisms, including learning, negative reinforcement, vulnerability, social context, and the capacity to regulate use. Reward-circuit involvement does not mean that every strongly rewarding activity or frequent habit is an addiction.
Craving can persist even when consumption is no longer particularly enjoyable, which motivates the distinction between incentive wanting and liking. Other theoretical traditions emphasize changes in affective opponent processes, habit systems, and prefrontal control; these models address partially overlapping data and remain areas of research. The 2025 Annual Review by Robinson and Berridge discusses such alternatives. Diagnosis, treatment planning, and prognosis require clinical assessment, not a claim that a scan shows an overactive reward center.
The popular analogy that social media notifications, sugar, exercise, gambling, and drugs are 'the same because dopamine' is scientifically inaccurate. Effects differ in pharmacology, intensity, environmental contingencies, harm, dependence potential, and individual vulnerabilities. For related food-specific interpretation, see Sugar and Dopamine: Reward, Motivation, and Common Myths and Is Sugar Addictive? What Human and Animal Evidence Actually Shows.
Reward processing, anhedonia, and mental health
Reward-related behavior can change during depression and several other mental health conditions. Researchers distinguish anticipatory interest, the ability to expend effort, learning from outcomes, and pleasure when rewards occur. The clinical word anhedonia refers to diminished interest or pleasure within an appropriate diagnostic context, but a single laboratory measure of reduced ventral striatal activation cannot confirm anhedonia or a specific disorder in an individual.
Medication experiments and neuroimaging may identify group-level differences and candidate mechanisms, but their diagnostic specificity, causal interpretation, individual predictive accuracy, and treatment relevance require separate validation. In particular, the 2025 pharmacological meta-analysis examined healthy adult participants, not a clinical diagnostic cohort. Its authors discussed potential future mechanistic biomarkers; this does not make reinforcement-learning task scores established tests for selecting psychiatric treatment.
Readers seeking clinical discussion can consult Anhedonia in Depression: Loss of Interest, Pleasure, and Motivation and Depression and the Brain: What Neuroscience Actually Shows. Changes in enjoyment, drive, concentration, or interest that are persistent and impairing deserve an assessment attentive to physical health, medication effects, sleep, stress, and psychological and social circumstances.
How scientists investigate brain reward circuits
Behavioral tasks and computational modeling
Researchers use probabilistic learning, monetary incentive delay, effort-discounting, reward-devaluation, choice, and taste-reactivity tasks to operationalize particular components of reward. Reaction time, choice proportion, effort expenditure, subjective ratings, and physiological measures quantify different outcomes. Computational models estimate latent variables such as expected value, learning rate, or prediction error, but model parameters are estimates contingent on assumptions and can be difficult to identify uniquely.
Functional MRI (fMRI)
Functional MRI commonly measures the blood-oxygen-level-dependent (BOLD) response, a vascular signal related indirectly to local neural activity. It offers useful spatial coverage of human reward circuitry and can compare responses during anticipation, decisions, and feedback. It does not directly count dopamine molecules, measure pleasure, or establish causality. A ventral-striatal BOLD difference can reflect multiple cognitive and physiological processes, including attention, action preparation, outcome learning, or vascular variation.
PET and molecular imaging
Positron emission tomography uses radiotracers to investigate aspects of receptor availability, transporter binding, metabolism, or changes in binding under particular conditions. Dopamine-related PET paradigms can be informative, but interpretation depends on tracer kinetics, baseline measures, competing neurotransmitter effects, and analytical assumptions. A PET image is not a universal snapshot of total brain dopamine or an individual motivation meter.
Electrophysiology and direct measurement
Scalp EEG records voltage differences generated by summed electrical activity and offers millisecond-scale timing, but limited specificity for deep reward structures. Intracranial human recordings and single-unit animal electrophysiology can resolve neural responses more directly in carefully defined circumstances. Fiber photometry and voltammetry are used in experimental animals to study neuronal-population or local chemical signals, with characteristic limits in specificity, sampling, and interpretation.
Causal manipulation and translation
Lesion studies, pharmacological interventions, direct electrical stimulation, and optogenetics can test causal hypotheses under controlled conditions. Optogenetics can target genetically defined populations in experimental animals; its findings cannot be treated as direct confirmation of human subjective experience. In humans, interventions often affect multiple brain and bodily systems, and observed behavior may have several explanations. Converging results from complementary methods are stronger than an isolated activation map.
Reverse inference is a recurring methodological error: observing activity in a region during a task and concluding that one particular mental state must have occurred. Many structures participate in multiple functions. Poldrack (2006) analyzed why anatomical selectivity and prior evidence matter when drawing inferences from brain-imaging data.
What is established, what remains unsettled?
Established evidence supports the central role of interconnected midbrain, striatal, pallidal, and prefrontal circuits in reward-related learning and motivated action; the existence of phasic dopamine patterns consistent with prediction errors in defined paradigms; the separability of some motivational and hedonic processes; and the influence of cost, context, and learning history on reward-guided behavior. Multiple experimental approaches contribute to these conclusions.
Active debates concern how diverse dopamine signals should be formalized, how reward identity and causal structure are represented, which circuit computations generalize across species and task designs, and how neural variables relate to complex human motivation outside laboratory conditions. Some interventions reveal causal mechanisms in animals but have narrower implications for human experience. Recent theoretical syntheses increasingly question whether any one-dimensional dopamine-as-value account covers all observations; see Kahnt and Schoenbaum (2025) and Gershman and colleagues (2024).
Clinical reward biomarkers remain a research aim. For a test to support individual diagnosis or treatment decisions, investigators must show adequate reliability, specificity, validity, calibration, and usefulness in the intended clinical population. A statistically significant association between a brain signal and symptoms does not meet those standards by itself.
Common myths about the brain reward system
Myth: Dopamine equals pleasure. Evidence instead implicates dopamine in learning, motivation, effort, and multiple forms of neural modulation; pleasure involves broader systems. The claim that every positive feeling is caused by a measurable dopamine 'hit' combines distinct processes.
Myth: The nucleus accumbens is the only reward center. It is an important hub in recurrent networks that include midbrain, cortex, pallidum, thalamus, hippocampus, amygdala, and other systems. Its subregions and neuronal populations also perform different functions.
Myth: More reward-system activity always means greater happiness. Neural activity can accompany expectation, surprise, salience, action, negative prediction error, or learned associations. Human happiness depends on multiple psychological and social processes and cannot be read directly from one brain region.
Myth: Every enjoyable product or habit is biologically addictive. Drug pharmacology, context, health consequences, repetitive behavior, and clinical diagnostic criteria differ profoundly. Dopamine participation does not establish an addictive disorder.
Myth: A scan can reveal whether somebody is lazy, unmotivated, or dopamine deficient. There is no general-purpose routine clinical brain scan that diagnoses these personal characteristics. Behavior is affected by opportunity, health, effort costs, social conditions, goals, and many other factors.
Myth: Reward prediction error explains every decision. It is a powerful family of models, but decisions involve inference, attention, context, planning, affect, and information-seeking, sometimes with multiple competing computational accounts.
Practical implications for learning, behavior, and daily life
The most useful general implication is that consequences, expectations, cues, and available alternatives shape how behavior changes. Predictable environments can make goals easier to pursue; feedback can supply information about progress; reducing unnecessary effort can change which option is selected. However, reward-circuit findings do not prescribe a universal schedule of rewards or justify claims about an ideal dopamine level.
Motivation can vary when sleep, hunger, health, emotional load, or competing demands change. A person's response to a reward also depends on meaning, autonomy, relationships, and prior experience. Practical guidance is better derived from evidence on behavior change and environmental design than from unvalidated dopamine-hacking routines. The English Hub analyses Motivation, Habits, and Self-Control and The Habit Loop: Cue, Routine, Reward, and Scientific Limits cover the behavioral applications.
Online advice about 'dopamine detox' often bundles reasonable choices—reducing distractions, resting, changing cue exposure—with inaccurate neurochemical explanations. Changing one's media habits can be useful, but temporarily avoiding pleasurable experiences should not be described as proven to reset dopamine receptors or cure addiction. Likewise, pharmacological and neurostimulation treatments must be evaluated in their specific medical contexts, with validated indications and professional oversight.
Frequently asked questions
What part of the brain controls reward?
Reward processing emerges from connected circuits, especially the ventral striatum, midbrain dopamine systems, prefrontal cortex, and ventral pallidum, with important input from amygdala, hippocampus, thalamus, and other regions. Their contributions vary by task and context.
Is the reward system the same as the dopamine system?
No. Dopamine systems influence reward processing and other functions, while reward-related circuits use several neurotransmitters and include areas that perform multiple kinds of computation. The systems overlap without being anatomically or functionally identical.
Why do unexpected rewards feel especially powerful?
Unexpected rewards can generate stronger learning signals because the outcome differs from what was predicted. That surprise may alter attention and future behavior; whether an event feels pleasurable depends on additional mechanisms and circumstances.
Can the reward system work without pleasure?
Some processes associated with reinforcement, incentive motivation, and action selection can operate with limited or altered hedonic experience. Clinical and experimental evidence supports partial dissociations between motivation and pleasure, though the terms require careful operational definitions.
How do habits relate to the reward circuit?
Consequences shape action learning, while repetition and contextual cueing may increase automaticity through interacting striatal and cortical systems. A habit need not be consciously rewarding each time it occurs, and changing a habit requires attention to environmental cues and alternative actions.
Does sugar activate the same reward circuits as addictive drugs?
Foods and substances can recruit some overlapping neural pathways, but their chemistry, potency, effects on physiology, learning contexts, and clinical implications differ. Shared circuit engagement does not prove equivalent addiction potential or diagnostic status.
Can stress change reward processing?
Stress can affect attention, motivation, effort, physiological state, and learning. Its effects depend on intensity, duration, context, and individual factors, so a universal claim that stress always increases or decreases dopamine and reward sensitivity is unwarranted.
Can doctors diagnose low dopamine from an fMRI scan?
Routine fMRI cannot quantify a person's brain dopamine concentration or diagnose a general dopamine deficiency. Specific medical disorders may warrant specialized neurological investigation, but a reward-task activation map is not an established stand-alone diagnostic test for low motivation or depression.
Does dopamine increase when people are motivated?
Motivated behavior often involves dopamine-sensitive mechanisms, but motivation is multidimensional and cannot be indexed by one chemical measurement. Dopamine manipulations may alter willingness to work, reward learning, response vigor, or choice differently depending on the situation.
How is reward prediction error different from reward itself?
Reward describes an outcome; reward prediction error describes the difference between that outcome and the prediction of it. A desirable outcome that is fully expected can have low modeled prediction error even when it is still enjoyable or valuable.
Related Articles
Explore the psychological science of rewards and motivation, approach and avoidance motivation, habit formation, and how the habit loop is interpreted in research.
For applied and clinical boundaries, see sugar and reward learning, coffee and dopamine, anhedonia in depression, and depression-related neuroscience.
References
Berridge, K. C., & Kringelbach, M. L. (2015). Pleasure systems in the brain. Neuron, 86(3), 646–664. https://doi.org/10.1016/j.neuron.2015.02.018
Berridge, K. C., & Robinson, T. E. (2016). Liking, wanting, and the incentive-sensitization theory of addiction. American Psychologist, 71(8), 670–679. https://doi.org/10.1037/amp0000059
Berridge, K. C., Robinson, T. E., & Aldridge, J. W. (2009). Dissecting components of reward: 'Liking,' 'wanting,' and learning. Current Opinion in Pharmacology, 9(1), 65–73. https://doi.org/10.1016/j.coph.2008.12.014
Gershman, S. J., et al. (2024). Explaining dopamine through prediction errors and beyond. Nature Neuroscience, 27, 1645–1655. https://doi.org/10.1038/s41593-024-01705-4
Haber, S. N., & Knutson, B. (2010). The reward circuit: Linking primate anatomy and human imaging. Neuropsychopharmacology, 35, 4–26. https://doi.org/10.1038/npp.2009.129
Kahnt, T., & Schoenbaum, G. (2025). The curious case of dopaminergic prediction errors and learning associative information beyond value. Nature Reviews Neuroscience, 26, 169–178. https://doi.org/10.1038/s41583-024-00898-8
Mahler, S. V., Smith, K. S., & Berridge, K. C. (2007). Endocannabinoid hedonic hotspot for sensory pleasure: Anandamide in nucleus accumbens shell enhances 'liking' of a sweet reward. Neuropsychopharmacology, 32, 2267–2278. https://doi.org/10.1038/sj.npp.1301376
Mkrtchian, A., Qiu, Z., Abir, Y., et al. (2025). Differential associations of dopamine and serotonin with reward and punishment processes in humans: A systematic review and meta-analysis. JAMA Psychiatry, 82(8), 818–829. https://doi.org/10.1001/jamapsychiatry.2025.0839
Poldrack, R. A. (2006). Can cognitive processes be inferred from neuroimaging data? Trends in Cognitive Sciences, 10(2), 59–63. https://doi.org/10.1016/j.tics.2005.12.004
Robinson, T. E., & Berridge, K. C. (2025). The incentive-sensitization theory of addiction 30 years on. Annual Review of Psychology, 76, 29–58. https://doi.org/10.1146/annurev-psych-011624-024031
Rutledge, R. B., Skandali, N., Dayan, P., & Dolan, R. J. (2014). A computational and neural model of momentary subjective well-being. Proceedings of the National Academy of Sciences, 111(33), 12252–12257. https://doi.org/10.1073/pnas.1407535111
Salamone, J. D., & Correa, M. (2024). The neurobiology of activational aspects of motivation: Exertion of effort, effort-based decision making, and the role of dopamine. Annual Review of Psychology, 75, 1–32. https://doi.org/10.1146/annurev-psych-020223-012208
Schultz, W. (2016). Dopamine reward prediction-error signalling: A two-component response. Nature Reviews Neuroscience, 17, 183–195. https://doi.org/10.1038/nrn.2015.26
Schultz, W., Dayan, P., & Montague, P. R. (1997). A neural substrate of prediction and reward. Science, 275(5306), 1593–1599. https://doi.org/10.1126/science.275.5306.1593