Does Drawing Actually Reduce Stress? What the Research Shows

Short answer: there is real evidence that making art lowers cortisol, and good evidence that being bad at it doesn't diminish the benefit, but the studies are small, and the effect was measured across roughly 45 minutes, not thirty seconds. What follows is what has actually been shown, caveats attached, because the caveats are the whole story.

The Cortisol Study

The most-cited work is Kaimal, Ray and Muniz (2016), published in Art Therapy. Thirty-nine healthy adults aged 18 to 59 spent 45 minutes making art from markers and paper, modelling clay, or collage materials, with no instructions about what to make. Saliva samples taken before and after measured cortisol, a hormone associated with stress.

Cortisol dropped by a statistically significant amount. Around 75% of participants finished with lower levels than they started with.

The detail worth pausing on: just under half the participants described themselves as having little art experience, and there was no correlation between prior experience and the size of the effect. Skill didn't matter. Neither did the medium. Clay worked about as well as markers.

The Caveats, Which Most Articles Skip

A small pilot study, and it deserves to be read as one.

About a quarter of participants went up, not down. Kaimal notes that some cortisol is necessary for normal functioning, so a rise isn't automatically bad, but it does mean the result wasn't universal.

There was no control group. Participants were measured before and after, with nothing to compare against.

Cortisol follows a daily rhythm. Levels run naturally highest in the morning and decline across the day. With a pre/post design and no control, you cannot cleanly separate the effect of art making from the decline that would have occurred anyway over 45 minutes.

Thirty-nine people. A pilot, and the authors present it as one.

None of this renders the finding worthless. It makes it a promising early result rather than a settled fact. The honest word is "suggestive," not "proven."

What Else Has Been Found

A related study from the same group found that roughly 45 minutes of open art making raised participants' self-efficacy, their confidence in their own capacity to handle a task. Different measure, similar direction, similarly small.

There is also emerging work using EEG and heart-rate measures during structured drawing, tracking how neural and autonomic activity shift over a session. Early, and a genuinely interesting direction, but not yet a body of evidence you can lean on.

What the Research Does Not Support

Bluntly, because this is a topic where marketing routinely outruns the evidence.

• It doesn't support a "10-second scribble" as a stress intervention. The measured sessions ran 45 minutes.
• It doesn't show drawing treats anxiety or depression. These are stress-marker and mood studies in healthy adults, not clinical trials.
• It doesn't establish drawing as superior to other activities. Nobody ran it against walking, knitting, or sitting quietly for 45 minutes.
• It doesn't show anything about brain waves proving one activity superior to another. Claims of that shape, applied to scrolling versus prompting versus drawing, have no research behind them at present.

If you are dealing with persistent anxiety or low mood, this is a pleasant thing to do and not a substitute for talking to a doctor or therapist.

What You Can Reasonably Take From It

• Making something with your hands for a decent stretch is plausibly good for your stress levels.
• You don't need to be good at it. That is the most robust finding in the set.
• The medium appears to matter little.
• Forty-five minutes is the duration actually tested. Shorter might work. Nobody has shown it.

A Note From Us

We make a drawing app, so treat our interest as declared. We would rather report the caveats than the headline. The cortisol result gets quoted constantly without the sample size, the missing control group, or the quarter of participants who moved the other way.

Draw something →

Frequently Asked Questions