Why Does Exercise Improve Your Mood? What's Actually Happening in Your Brain

Ask most people why exercise puts them in a better mood and you’ll get some version of “endorphins.” That answer isn’t wrong exactly, but it’s also not where most of the actual science points anymore. The real picture involves several overlapping systems: neurotransmitters, a growth factor called BDNF, a lesser known signaling system called the endocannabinoids, a stress hormone axis, and the immune system. None of these work alone, and none of them fully explain the effect by themselves. Let’s go through what’s actually happening.

Your brain chemistry shifts almost immediately

During exercise, your body pulls branched-chain amino acids into working muscle. That frees up more tryptophan to cross into the brain, where it gets converted into serotonin [1]. At the same time, dopamine synthesis and release increase in reward-related regions like the nucleus accumbens and prefrontal cortex, and norepinephrine rises in the brainstem and cortex in proportion to exercise intensity [1]. Together these three changes touch mood, motivation, and alertness all at once, which is part of why a workout can feel like it’s doing several things to you simultaneously.

It’s worth noting that this isn’t just “more serotonin equals better mood.” Recent umbrella reviews have pushed back hard on the old idea that depression is simply caused by low serotonin, and there’s no strong evidence that people with depression have reduced baseline serotonin activity [1]. What exercise seems to do instead is recalibrate several neuromodulatory systems together: serotonin, dopamine, norepinephrine, and GABA all shift in a coordinated way, and that coordination looks more important than any single chemical rising or falling [1].

Dopamine’s role deserves its own mention, because it’s often oversimplified as “the pleasure chemical.” A recent review of the exercise-dopamine relationship makes the case that dopamine’s real function during exercise is less about pleasure and more about reward learning and motivational salience, meaning it helps your brain assign value to the activity itself so you’re inclined to do it again [4]. Aerobic exercise and HIIT tend to broadly activate reward processing and BDNF expression, while resistance training seems to have a more targeted effect on dopamine receptor availability in the nucleus accumbens [4]. These aren’t interchangeable effects, they’re distinct biological signatures depending on what kind of exercise you do.

BDNF: the part that builds rather than just borrows

One of the more consistent findings across the literature is that exercise increases brain-derived neurotrophic factor, or BDNF, a protein that supports neuron growth, survival, and the strengthening of synaptic connections [1]. Even a single 15-minute bout of moderate aerobic exercise can raise peripheral BDNF levels, and the increase typically peaks 10 to 15 minutes after exercise ends [2]. People with mood disorders tend to have lower baseline BDNF, and levels rise with exercise in a way that’s been correlated with symptom improvement [1].

What’s interesting is how BDNF gets there. Skeletal muscle releases signaling proteins called myokines during exercise, and one of them, cathepsin B, crosses the blood-brain barrier and directly increases BDNF expression [1]. Lactate, which builds up during exercise depending on intensity, also appears to trigger BDNF release even when infused without any actual physical activity [1]. So part of what’s happening is literally your muscles sending a chemical signal to your brain that says “build more capacity.”

This matters for cognition too, not just mood. BDNF supports long-term potentiation, the process underlying memory formation, and one review found that a year of aerobic exercise prevented the decline in hippocampal volume that typically comes with aging [1]. The mood and cognitive benefits aren’t really separate effects, they’re downstream of the same growth signal.

The chemical behind the “runner’s high”

For decades, the euphoric feeling some people get from a long run was attributed to endorphins. The problem is that endorphins are hydrophilic peptides that don’t cross the blood-brain barrier well, which raises real doubts about whether they can directly cause a centrally felt euphoria [2]. Several studies have blocked the opioid system entirely with naltrexone and still found comparable euphoria and anxiety reduction after running [2].

The stronger candidate now is the endocannabinoid system, specifically anandamide (AEA). Unlike endorphins, endocannabinoids are lipophilic and cross the blood-brain barrier easily [3]. Circulating AEA reliably increases during moderate-intensity endurance exercise, and this rise correlates with the anxiolytic and euphoric features people describe as the runner’s high [2,3]. A study comparing humans and dogs, both endurance-adapted species, to ferrets, which aren’t built for sustained running, found that only the endurance-adapted species showed this AEA increase, suggesting the whole system may have evolved specifically to make long-distance movement feel rewarding [3].

There’s a fairly narrow window where this effect actually shows up. AEA production is highest below the anaerobic threshold, running produces a stronger effect than cycling, and the exercise needs to run at least 20 minutes, with peak mood benefits usually appearing after 30 to 35 minutes [2]. Go too easy and you won’t trigger much of a response. Go into pure high-intensity anaerobic territory and the response shifts too.

Cortisol isn’t the villain here

It’s tempting to think of cortisol purely as “the stress hormone” and assume exercise-induced cortisol spikes are bad news. That’s not quite right. Aerobic exercise activates the hypothalamic-pituitary-adrenal (HPA) axis and does raise cortisol acutely, but this cortisol spike actually triggers the synthesis of endocannabinoids within the hypothalamus, which then provide negative feedback that dampens the stress response back down [3]. Regular training seems to improve this feedback loop over time, producing lower resting cortisol, better glucocorticoid receptor sensitivity, and a more efficient stress response overall [1].

The catch is intensity and recovery. When training volume chronically outpaces recovery, this same system can flip into something that looks more like chronic stress: persistently elevated cortisol, blunted hippocampal neurogenesis, and mood disturbance [1]. The HPA axis response to exercise follows a dose-dependent curve rather than a straight line, meaning more exercise doesn’t linearly translate into more benefit.

Inflammation gets quieted down too

Chronic low-grade inflammation, particularly elevated interleukin-6, TNF-alpha, and IL-1 beta, has been linked to worse outcomes in depression and anxiety through its interference with neurotransmitter function and BDNF signaling [1]. Regular moderate exercise reduces baseline levels of these inflammatory markers, increases anti-inflammatory cytokines like IL-10, and improves vagal tone, which is itself associated with better mood regulation [1].

Here too, intensity matters. A single vigorous session actually causes a short-term inflammatory spike, but this is thought to function as an adaptive signal that muscle tissue uses to mobilize energy and kick off downstream anti-inflammatory processes [1]. It’s the chronic, well-recovered pattern of exercise that produces the net anti-inflammatory effect, not any single hard session.

So what’s the actual dose?

Across the literature, a fairly consistent picture emerges. Single sessions of 20 to 60 minutes at moderate intensity reliably reduce negative mood and boost positive affect [1]. For longer term change, most reviews point to an 8 to 12 week window of 3 to 5 sessions weekly as the sweet spot, and interestingly, benefits don’t scale endlessly with intensity. Low-intensity programs have produced improvements comparable to higher intensity ones in several analyses, and some reviews even found effects plateauing or declining once programs extended past 12 weeks [1].

Takeaway

There isn’t one mechanism that explains why exercise improves mood. It’s closer to several systems moving in the same direction at once: neurotransmitters shift to support alertness and reward, BDNF gets released to build new neural capacity, endocannabinoids create a genuine euphoric and anxiety-reducing state, cortisol gets recalibrated rather than simply spiked, and inflammation quiets down over time. None of these require anything more complicated than a brisk walk or a jog done consistently. The biology is doing the work, you just have to show up for it.

Common Questions

Does exercise really work as well as antidepressants for mood?

For depressive and anxiety disorders, exercise has produced effect sizes comparable to therapy and pharmaceutical interventions in several meta-analyses, and it comes with added physical health benefits those treatments don't provide. It's generally framed as an effective adjunct rather than a replacement.

How much exercise do I actually need to feel a mood benefit?

A single session of 20 to 60 minutes at moderate intensity is enough to produce measurable mood improvements. For longer term benefits, research most consistently points to 8 to 12 weeks of 3 to 5 sessions per week.

Is the 'runner's high' just endorphins?

Not primarily. Endorphins don't cross the blood-brain barrier well, so most current research points to endocannabinoids, particularly anandamide, as the main driver of the euphoric, anxiety-reducing state people feel after a run.

Can exercise be too intense to help your mood?

Yes. The relationship follows a dose-dependent curve. Moderate, well-recovered training produces the clearest benefits, while chronic overtraining without adequate recovery can blunt or even reverse some of these effects.

References