The Runner's High: What Actually Causes It (Hint: Not Endorphins)

Educational journalism, not medical advice. Every claim here is checked against its cited sources by editor Tim Bunce — a health writer, not a physician. It isn’t specific to your situation: for health decisions, talk to your own clinician. How we work →

Why “endorphins” was the wrong answer

The endorphin hypothesis traces to the late 1970s, when researchers measured peripheral beta-endorphin levels in runners’ blood after intense exercise and observed elevations. The leap from “blood endorphins go up” to “the brain feels euphoric because of endorphins” was popular but mechanically suspect. Beta-endorphin is a large peptide (~3.5 kDa); the blood-brain barrier excludes peptides of that size. Peripheral endorphin doesn’t reach the brain’s opioid receptors.

The situation got murkier with the 2008 Boecker et al. PET study, which used [11C]diprenorphine (a brain-penetrant opioid radioligand) to image opioid binding in 10 athletes before and after a 2-hour run. They found significantly increased opioid binding in frontal and limbic regions, and the magnitude correlated with euphoria self-reports Boecker 2008. So the central opioid system is activated — just not by the peripheral endorphins everyone was measuring.

But opioid activation isn’t the whole story.

The endocannabinoid pivot

Endocannabinoids (anandamide, 2-AG) are lipid signalling molecules that do cross the blood-brain barrier. They bind the same CB1 and CB2 receptors that THC binds. Beginning in the early 2000s, exercise studies started measuring peripheral endocannabinoids after running and found consistent elevations.

The decisive experiment came in 2015. Fuss et al. ran mice for prolonged voluntary wheel-running and tested for two runner’s-high markers (reduced anxiety in elevated-plus-maze, reduced thermal pain sensitivity). They then administered:

• Naloxone (opioid receptor blocker): runner’s-high effects mostly preserved.
• AM251 (CB1 cannabinoid receptor blocker): runner’s-high effects abolished.
• Both blocked: complete abolition.

The conclusion: the dominant runner’s-high mechanism in mammals is endocannabinoid-mediated, with opioid activation as a secondary contributor Fuss 2015.

The 2012 Raichlen et al. study extended the human data: 12 college athletes ran on a treadmill at three intensities; only the moderate-intensity condition (~70–75% HR-max) produced significant peripheral anandamide elevation, suggesting an intensity sweet spot Raichlen 2012.

“Wheel running induces a runner’s high in mice, depending on cannabinoid type 1 (CB1) receptors on GABAergic neurons in the limbic system. Peripheral and central measurements indicate that endocannabinoid signaling, not opioid signaling, is necessary for sustained running-induced reductions in anxiety and pain sensitivity.”

— Fuss et al., PNAS, 2015 view source

The evolutionary story

Why would mammals have evolved a neurochemical reward for sustained running? The 2012 Raichlen et al. comparative study tested ferrets (non-cursorial) and dogs and humans (cursorial mammals); only the cursorial species elevated endocannabinoids during exercise. The implication: the runner’s high may be an evolved adaptation that supported persistence hunting and long-distance migration in our hunter-gatherer ancestors Raichlen 2012.

This is speculative but consistent with multiple lines of evidence:

• The phenomenon is universal across mammalian cursorial species but absent in sprinters.
• The intensity sweet spot (~70–75% HR-max) overlaps with the metabolic zone for sustained aerobic locomotion.
• The duration threshold (typically >30 minutes of continuous running) matches what would have been required for ancient persistence hunting bouts.

Intensity and duration sweet spot

Practical findings from the human studies:

• Intensity: roughly 70–75% of max heart rate. Easy jogging is too low; sprint intervals are too high. The Raichlen 2012 data show no anandamide elevation at low-intensity walking and reduced (vs moderate) elevation at very-high-intensity sprinting.
• Duration: at least 30 minutes of continuous activity. Most reported runner’s-high experiences happen in the 45–90-minute range. The phenomenon doesn’t reliably appear in 20-minute sessions.
• Continuity: the literature is less clear on intervals vs steady-state. Some studies suggest interval running interrupts the build-up; observational reports are mixed.
• Modality: cycling and swimming can produce comparable effects in trained subjects but at higher intensities/durations.

Why some people get it and others don’t

The runner’s high is not universal. Survey data (Markoff 1982 and later replications) consistently show 30–50% of runners report rarely or never experiencing classic euphoric runner’s high. Variables that predict reporting:

• Training status: more-trained runners report the experience more frequently. Aerobic fitness allows sustained moderate-intensity running long enough to reach the threshold.
• Genetic CB1 receptor sensitivity: emerging genome-wide work suggests CB1 polymorphisms may explain part of the variance.
• Pace control: runners who consistently run at the wrong intensity (too easy or too hard) may not hit the sweet spot.
• Attention and expectation: subjects primed to expect a runner’s high report it more often, suggesting some attentional/placebo overlay on top of the neurochemistry.

Other contributing systems

Endocannabinoids and opioids aren’t the whole story. Other systems that activate during sustained running and contribute to mood/cognitive effects:

• BDNF (brain-derived neurotrophic factor): elevated during and after exercise; supports neurogenesis and synaptic plasticity. The 2014 Sleiman et al. studies on BDNF and exercise suggest it’s upstream of much of the chronic mood benefit of running, even if it’s not the acute “high”.
• Dopamine: chronic exercise upregulates dopaminergic signalling. Acute effects on euphoria are smaller than for endocannabinoids.
• Serotonin: exercise modestly affects serotonergic activity; the chronic antidepressant effect of running likely involves this system more than endorphin/endocannabinoid surges.
• Adrenocorticotropic hormone (ACTH) and cortisol: increase during stress and exercise, with complex downstream interactions on mood.

Can you train the runner’s high?

Partly. The factors under your control:

• Build the aerobic capacity to run at 70–75% HR-max for 45+ minutes. If your easy pace already approaches that intensity, you’re running “too hard” for the sweet spot.
• Use a heart rate monitor or pace by perceived exertion to land in the sweet spot consistently.
• Don’t fixate on the experience — expectation can both prime and inhibit it.
• Try varied terrain and aesthetic environments; some studies suggest greenspace running produces the experience more readily than treadmill.

What you can’t train: the underlying CB1 receptor density and sensitivity. Some people will get the experience more readily than others for genetic reasons.

Common myths

• “Endorphins cause the runner’s high.” Mostly wrong. Peripheral endorphins don’t cross the blood-brain barrier; central opioid activation is real but secondary to endocannabinoid signalling.
• “You just need to run far enough.” Distance alone doesn’t guarantee it. Intensity matters more — running too easy or too hard misses the sweet spot.
• “The runner’s high is a myth.” Also wrong. The neurochemistry is well-documented in mice and humans. Individual variability is large but the phenomenon is real.
• “If you don’t feel a runner’s high, you’re not running enough.” Not necessarily. ~30–50% of runners rarely experience it, often for genetic and intensity-control reasons. The chronic mood and health benefits of running don’t require the acute “high”.

Practical takeaways

• The runner’s high is primarily endocannabinoid-mediated, not endorphin-mediated.
• Sweet spot intensity: ~70–75% of max heart rate. Sweet spot duration: 30+ minutes of continuous moderate-intensity running.
• 30–50% of runners rarely or never report the classic euphoric experience — this is normal individual variation.
• The experience is distinct from general exercise mood lift, which is more reliable and probably mediated by BDNF and serotonin over chronic time scales.
• You can train the conditions (fitness, pacing) but not the underlying receptor biology.
• Don’t fixate on chasing it — the chronic mood and health benefits of running don’t require it.

What the human evidence actually shows

Most of the headline-grabbing endocannabinoid work was done in mice, and that is worth being honest about. The famous experiment that toppled the endorphin story used genetically modified rodents, not people, so the obvious question is whether the same machinery runs in a human brain. The strongest answer so far comes from a controlled trial in people. Researchers had 63 adults run on a treadmill for 45 minutes at a moderate pace on one day and walk on another, then measured both their mood and their blood chemistry. Running raised self-reported euphoria, lowered anxiety, and increased circulating levels of the two main endocannabinoids, anandamide (AEA) and 2-arachidonoylglycerol (2-AG). Crucially, when participants were given naltrexone, a drug that blocks opioid receptors, the high still appeared and the endocannabinoid levels still rose Siebers 2021. In other words, switching off the opioid system in humans did not switch off the runner's high, which is exactly what you would expect if endocannabinoids, not endorphins, are doing the heavy lifting.

That single trial does not stand alone. A 2022 systematic review and meta-analysis pooled the available studies and found a consistent rise in both anandamide and 2-AG after a single bout of exercise, across different activities (running and cycling among them) and across both humans and animals; roughly three-quarters of the human samples that measured anandamide showed a significant increase after acute exercise Desai 2022. The same review flagged a lot of variability from study to study, which it attributed to differences in exercise intensity, fitness, when the blood was drawn, and whether people had eaten beforehand. The practical upshot is that the endocannabinoid rise is real and reproducible in people, but its size is sensitive to how the workout is done and how it is measured. It is also a useful corrective to the idea that the runner's high is a running-only phenomenon: the chemistry shows up in cyclists too, so the "runner's" label is more about history than about anything unique to running.

It is fair to ask how settled all of this is. The honest position, shared by the researchers themselves, is that the human evidence makes endocannabinoids "strong candidates" rather than a proven sole cause Siebers 2021. Blood measurements are taken in the periphery, not directly in the brain, and a single drug-blockade trial in dozens of people is not the same as decades of replication. The story is more solid than the endorphin myth it replaced, but it is still an active area of research rather than a closed case.

The mood payoff that lasts longer than the high

For most readers the runner's high is interesting mainly because it hints at something larger: exercise reliably lifts mood, and the endocannabinoid system appears to be one of the levers. The clearest demonstration that this is more than a fleeting buzz comes from work on fear and anxiety. In a randomized controlled trial, 35 women with post-traumatic stress disorder completed a fear-extinction protocol and were then assigned to either 30 minutes of moderate-intensity exercise (70 to 75 percent of maximum heart rate) or a light-intensity control. The exercise group showed lower threat responses the next day, and statistical analysis indicated that the rise in anandamide (along with the growth factor BDNF) was part of what carried that benefit Crombie 2021. The same anandamide surge that contributes to the in-the-moment euphoria, in other words, may also help the brain re-learn that a once-threatening cue is now safe. That is a preliminary finding in a small, specific group, so it should be read as a promising mechanism rather than a treatment, but it links the high to something genuinely useful.

Step back from the single-session chemistry and the population-level picture is robust. A 2023 umbrella review in the British Journal of Sports Medicine synthesized 97 systematic reviews covering more than a thousand trials and over 128,000 participants. It found that physical activity produced medium-sized reductions in depression (median effect size -0.43), anxiety (median effect size -0.42), and psychological distress (effect size -0.60) compared with usual care, and the authors argued that exercise should be a mainstay in managing these conditions alongside therapy and medication Singh 2023. Two cautions keep this honest. First, the mood-lifting effect of regular exercise is bigger and better established than the runner's high itself, and it does not require ever feeling euphoric mid-run, so a reader who never gets a "high" is not missing the main benefit. Second, exercise is a complement to, not a replacement for, professional care; anyone managing depression, anxiety, or PTSD should talk to a clinician before changing a treatment plan.

When chasing the high becomes a problem

The flip side of a built-in reward system is that rewards can be over-pursued. A small minority of people develop what researchers call exercise dependence or exercise addiction: training that has become compulsive, continues despite injury, and crowds out work, relationships, and rest. A literature review of endurance athletes describes the same hallmarks used to define other behavioral addictions, including salience (exercise dominating a person's life), tolerance, withdrawal-like symptoms when a session is missed, and conflict with family or work, and notes that endurance sports carry among the highest risk Nogueira 2018. Reported prevalence estimates vary widely, from roughly 3 to 42 percent depending on the population and the questionnaire used, which mostly reflects how loosely the term has been defined; the high end comes from screening tools applied to committed athletes, not the general public Nogueira 2018.

The same review points out that the neurochemistry cuts both ways. The reward and endorphin systems that make exercise feel good are also among the proposed mechanisms for why it can become compulsive, and one cited study reported that people with exercise addiction had lower anandamide levels and that a two-week break worsened their mood Nogueira 2018. That is a single observation rather than settled science, but it is a useful reminder that "the high feels great" is not, by itself, evidence that more is always better. The practical signal to watch for is not how much someone runs but whether the running has stopped being a choice: training through a stress fracture, skipping social or family commitments to fit in mileage, or feeling anxious and irritable on rest days. Compulsive exercise also travels with disordered eating, particularly in endurance sport Nogueira 2018. If those patterns sound familiar, that is a conversation to have with a doctor or mental-health professional rather than something to push through.

Honest limits and open questions

It would be neat to say science has fully swapped endorphins out for endocannabinoids, but the truth is messier and more interesting. The brain-imaging study that originally fueled the opioid theory was not wrong about its data: it found that running reduced opioid-receptor availability in prefrontal and limbic brain regions and that the drop tracked with how euphoric the runners felt Boecker 2008. The endocannabinoid trials do not erase that finding; they show that blocking opioids does not abolish the high Siebers 2021. The most defensible reading is that the opioid and endocannabinoid systems are both engaged during hard exercise and probably interact, with endocannabinoids appearing necessary for the classic high and opioids likely contributing texture to the experience. "It's the endocannabinoids" is a better headline than "it's the endorphins," but "it's only one molecule" is unlikely to be the final word for any of the candidates.

Several genuine unknowns remain. Almost all the human endocannabinoid measurements come from blood, not the brain itself, so they are an indirect read on what receptors are doing centrally Desai 2022. The size of the endocannabinoid response varies a lot between studies and people, and we do not yet have a clean account of why some individuals reliably feel a high and others rarely do Desai 2022. And the evolutionary "born to run" framing, while appealing, is built on comparisons between species rather than direct proof that the high exists to reward persistence hunting Raichlen 2012. None of this should undercut the bottom line a reader can act on: moderate, sustained aerobic exercise reliably nudges mood-related brain chemistry in a good direction, the euphoric high is a real but optional bonus on top of a much more dependable improvement in mood, and the open questions are about precise mechanism, not about whether the effect is worth having.

References