Expressive writing — writing about difficult or emotionally significant experiences — has been studied since the 1980s, in a body of research more substantial than most self-help practices can claim.
The findings are moderate, specific, and not what most journaling advice describes.
The original paradigm
The standard protocol used in this research is quite specific: writing for around fifteen to twenty minutes, on three or four consecutive days, about a difficult or traumatic experience, exploring both the events and the feelings associated with them.
Participants are told not to worry about grammar or spelling, and that nobody will read it.
Control conditions typically involve writing about neutral topics for the same duration.
What has been found
Meta-analyses have reported small to moderate effects across various outcomes, including physical health measures, psychological wellbeing, and in some studies practical outcomes like reduced medical visits.
Effects have been reported on immune markers in some studies, and this literature is less consistent.
The effect sizes are modest. Enthusiastic descriptions of the research overstate them considerably, and the effects are real and have replicated across many studies.
Notably, benefits are frequently not immediate. Participants sometimes feel worse directly after writing and better weeks later, which complicates any evaluation based on how it feels at the time.
What appears to matter
Analysis of what distinguishes writing that produces benefit has identified some features.
Both events and emotions. Writing that describes what happened without addressing feelings, or that expresses feelings without describing events, appears less effective than writing that does both.
Change in perspective across sessions. Studies analysing text have found that participants who shift pronoun use — moving between first person and other perspectives — show greater benefit.
The proposed explanation is that shifting perspective indicates cognitive processing rather than repeated rehearsal of the same account.
Increasing use of causal and insight words. Text analysis has found that participants whose writing shows increasing use of words indicating explanation and understanding across sessions benefit more.
Which suggests the mechanism involves constructing a coherent account rather than simply venting.
Not writing about the same thing repeatedly without development. Rumination in written form appears not to help and may make things worse.
What doesn't appear to help
Pure venting. Expressing distress without any structuring or explanation. Some evidence suggests this can reinforce distress rather than reducing it.
Writing while acutely distressed. Studies generally involve writing about past experiences rather than ongoing crises, and the findings shouldn't be extended to acute situations.
Very short sessions. The protocol uses fifteen to twenty minutes, and briefer writing has less support.
Gratitude journaling
A separate practice with its own literature, frequently conflated with expressive writing.
Studies of gratitude interventions — listing things one is grateful for, regularly — have found effects on wellbeing measures, with meta-analyses reporting modest effects.
Some findings suggest that less frequent practice may be more effective than daily, possibly because daily listing produces habituation and the exercise becomes rote.
Effects on depression and anxiety specifically appear smaller than effects on general wellbeing measures.
The practical version
Combining what the research supports.
For processing something difficult: write about a specific past experience for fifteen to twenty minutes, on several consecutive days. Include what happened and how you felt about it. Attempt to explain rather than only describe. Expect to feel worse immediately and better later.
For general wellbeing: periodic rather than daily. Specific rather than general — particular things rather than categories. And varied, to avoid the same items appearing each time.
What to avoid: repeatedly writing the same account of an ongoing situation without any development, which is rumination with a pen.
The cautions
Some genuine ones.
Writing about trauma can be distressing, and for people with significant trauma history this is not a self-help exercise. Structured trauma-focused therapy exists and is considerably better supported for that population.
The research protocol involves writing that nobody reads. Journals that might be read by others produce different writing, and the privacy appears to matter.
And the effects, while real, are modest. Journaling is not a substitute for treatment where treatment is indicated, and framing it as one does a disservice to people who need more.
Why it's under-recommended
Worth noting that this is among the better-evidenced self-help practices and is promoted less than several with weaker support.
The likely reason is commercial. It requires paper and time, and nobody can sell it. Which is a pattern worth noticing generally when assessing what gets recommended.
General information only. If you are dealing with trauma or persistent low mood, please speak to a qualified professional.
Digital or paper
A question people ask and the research does not clearly answer, since most studies predate the alternative or used paper by default.
Some evidence on handwriting suggests it engages processes differently from typing, with studies on note-taking finding differences in retention and in how material is processed.
Whether that transfers to expressive writing is unestablished.
What may matter more is privacy. The research protocol involves writing that nobody reads, and a file on a synced device carries a background awareness of potential discovery that a notebook does not. Anything that makes you write more carefully is working against the exercise.