Journaling Grief, And What The Trials Actually Found

I read every page ranking for this and none of them mentions the one study a person considering this should probably know about. Not because it's damning. Because it's honest, and honest is easier to hold.

So, directly. Journaling grief is a reasonable thing to do and it is not a treatment for grief. The best-designed test of the standard emotional-writing task in bereaved people, run by Margaret Stroebe and colleagues in 2002, found no reduction in distress and no reduction in doctor visits, either immediately or six months later. That doesn't mean don't write. It means the reason to write isn't that it processes the loss faster. The prompts further down are built on a different model, the dual process model of coping with bereavement, which treats grieving as an alternation between facing the loss and rebuilding around it, and which gives a journal a job it can actually do.

Read This Part First

Not at the bottom. Here.

A journal is not treatment and I'm not going to imply otherwise anywhere on this page. If what you're carrying is heavier than writing can hold, a qualified professional is the right call, and in immediate crisis a crisis line is the right call and this article is not. In the US that's 988, callable, textable or chattable, running 24 hours a day, every day of the year, and it's explicitly for emotional distress and "just need someone to talk to," not only for people who are actively suicidal. Outside the US your national line will be one search away and it will be staffed tonight.

There's a second thing that belongs up here rather than in a footnote. The research on emotional writing has found harm as well as benefit. The clinical review by Karen Baikie and Kay Wilhelm reports that expressive writing "was detrimental for adult survivors of childhood abuse and for a small sample of eight Vietnam veterans with PTSD," and states plainly that it "should not replace appropriate medical or psychological treatment in clinical populations; it should be used as an adjunct to standard treatment" (Advances in Psychiatric Treatment, 2005). They also note that the usual immediate effect of a session is a short-term increase in distress and low mood.

If the death you're writing about was traumatic, sudden, violent, or is tangled up with something older, that's the exact territory where the writing research found people doing worse rather than better. Please do that part with someone, not alone with a notebook at one in the morning.

The NHS puts the general threshold about as plainly as anyone. Their guidance is to see a GP if "you're struggling to cope with stress, anxiety or a low mood," if "you've had a low mood for more than 2 weeks," or if "things you're trying yourself are not helping" (NHS on grief and bereavement). A journal is a thing you're trying yourself.

The Trial The Ranking Pages Don't Mention

Now the uncomfortable part, and I'd rather you hear it from someone who isn't selling a prompt kit.

Margaret Stroebe, Wolfgang Stroebe, Henk Schut, Emmanuelle Zech and Jan van den Bout published two studies together in the Journal of Consulting and Clinical Psychology in 2002, volume 70, pages 169 to 178 (abstract here). Study one followed bereaved people prospectively over two years and asked whether talking about the loss and disclosing emotions predicted better adjustment after losing a partner. Their reported finding is that there was "no evidence that disclosure facilitated adjustment."

Study two is the one that matters more, because it was randomised. Recently bereaved people were assigned either to the standard Pennebaker writing task or to a no-essay control. The writing task "did not result in a reduction of distress or of doctors visits either immediately after the bereavement or at a 6-month follow-up." And they checked the obvious escape hatches. No benefit for people who'd suffered an unexpected loss. No benefit for the ones who still reported a high need to disclose.

That's a null result in the population everyone assumes benefits most.

I don't read it as writing being useless after a death. I read it as the specific claim being wrong, the claim that structured emotional disclosure is a mechanism which speeds up bereavement. The broader picture for emotional writing across all populations is a small positive average, an r of .075 pooled across 146 randomised studies (Frattaroli, Psychological Bulletin, 2006), and grief looks like one of the places where the average doesn't show up.

Nobody in the top ten says this. One of them, a hospice-affiliated page with thirty-one prompts on it, doesn't cite any research at all and doesn't carry a crisis line either.

The Model That Gives A Journal A Job

If not processing, then what.

The framework I'd build a grief journal on is the dual process model of coping with bereavement, from Margaret Stroebe and Henk Schut, published in Death Studies in 1999 and revisited a decade later in Omega (the ten-year review is here). Instead of stages, it describes two kinds of stressor and a movement between them.

Loss-oriented coping is the part everyone pictures. A study analysing two national bereavement surveys describes it as "focussing on the loss of the deceased person (LO coping, e.g. grieving)" (Torrens-Burton et al., Palliative Care and Social Practice, 2022). Remembering, yearning, going through the photographs, missing them.

Restoration-oriented coping is the half that gets left out of prompt lists entirely. The same paper describes it as "negotiating the practical and psychosocial changes to their lives that occur as a result of the bereavement (RO coping, e.g. forming new roles/identities/relationships)." Learning the admin nobody warned you about. Working out who you are now.

And the mechanism connecting them is oscillation, the back-and-forth. The Loss Foundation, a UK bereavement charity, describes it as alternating "between confronting the pain of the loss and adapting to life changes after the loss," and notes it can look like crying one day and laughing the next (their overview of the model).

Here's why that matters for a notebook specifically. Every grief prompt list I found is entirely loss-oriented. Thirty-one prompts about the person, the memories, the missing. Seven prompts, same shape. If oscillation is the thing being described as adaptive, then a journal made of nothing but loss-oriented prompts is a tool that only turns one way.

Prompts For The Loss Side

These are the ordinary ones and they belong here. I've kept them concrete rather than sweeping, because "how has grief changed you" is a question for a year from now and not for a Tuesday.

• What happened in the last week that I wanted to tell them about?
• What's a small thing about them nobody else would remember?
• What did they get wrong, and am I allowed to write that down?
• Which object in this house is doing the most work right now?
• What did I not say, and would I have said it if I'd known?
• What's the version of them I'm at risk of smoothing out?
• Write it as a letter. Start with whatever you'd actually open with, not a greeting.
• What was ordinary about them that I'll lose first?

That last one has a reason behind it. The specific and unremarkable details go before the big ones do, and a journal is one of the only places they survive. Not as therapy. As a record.

If a blank page is the obstacle rather than the material, unfinished sentences tend to work better than questions, and there's a whole set of sentence-stem entry starters built for exactly that friction.

Prompts For The Restoration Side

This half feels wrong to write and that's sort of the point.

People report guilt around restoration-oriented material, as though attending to the practical rebuilding is a betrayal or a failure to grieve properly. In the model it isn't. It's half the process.

• What did I handle this week that I didn't know how to do a month ago?
• What has to be decided in the next fortnight, and who can I ask?
• Which routine has quietly collapsed, and does it need rebuilding or replacing?
• Who's still turning up, and have I told them?
• What am I now responsible for that used to be theirs?
• What's one hour of this week that was fine? Just describe it.
• What do I want the next six months to contain, in plain concrete terms?
• Which relationship needs a call from me rather than from them?

The sixth one is the hardest to write and I'd argue it's the one worth keeping. Naming an hour that was fine is not a betrayal of anyone. It's also the entry that's most useful to read back in a year, when the memory of this period will have flattened into one undifferentiated colour.

The Oscillation Rule

Here's the actual protocol, which is mine rather than the researchers', built from their model. Nobody in the search results does this and I'd want you to treat it as a structure rather than a finding.

Pair your entries. One loss-oriented, one restoration-oriented, in the same sitting or across two days. Never a run of four loss entries with nothing in between.

That bottom row is the whole design. Neither column is the healthy one. The alternation is the thing the model actually describes, and a journal is unusually good at enforcing an alternation because you can see on the page which side you've been living on.

One caution. If pairing feels like a rule you're failing at, drop it. The last thing a bereaved person needs is a productivity system with a compliance metric attached.

What The Prompt Lists Get Wrong

Four things, quickly, and I'll admit some of these are preference.

They assume writing is safe by default. The evidence says it usually is and sometimes isn't, and the sometimes clusters in exactly the losses people describe as traumatic.

They frame the journal as a process that ends. Prompts about "the lessons grief taught me" belong to a much later chapter and reading them in month one mostly produces a sense of being behind schedule.

They almost all skip restoration. See above.

And they treat frequency as a virtue. Nothing I found supports daily grief writing over occasional grief writing, and given that the immediate effect of a session is a documented short-term dip in mood, a daily cadence means a daily dip. I'd write when there's something to write, and I'd stop mid-sentence when I wanted to. You're allowed to close the notebook.

When It's Been A Long Time

This part needs care, so I'll be precise about what I'm saying and what I'm not.

Clinicians now have a diagnosis for grief that stays severe and disabling for a long stretch. It's called prolonged grief disorder, and it exists in both major diagnostic systems. In the DSM-5-TR the criterion is written as "the death, at least 12 months ago, of a person who was close to the bereaved." In ICD-11 the framing is different, with the note that "grief responses lasting for less than 6 months, and for longer periods in some cultural contexts, should not be regarded as meeting this requirement." Both are summarised in Maarten Eisma's review of the two definitions and the arguments around them (Australian and New Zealand Journal of Psychiatry, 2023).

I'm not telling you to diagnose yourself off a blog post, and those thresholds are entry conditions rather than the diagnosis, which also requires specific symptoms and significant impairment assessed by a clinician. The only reason it's on this page is that a lot of people quietly believe there's an expiry date on being allowed to still be wrecked, and there isn't, and the existence of a named condition with treatment attached is better news than the alternative. If it's been a long time and it hasn't shifted at all, that's a thing to take to a doctor. It isn't a thing to fix with a better prompt.

What I'd Say, With No Receipt To Offer

I'm not going to write a paragraph about my own loss, because that's not something I'm going to manufacture for a piece of search traffic, and you'd be able to tell.

What I'll offer instead is a judgement. Of everything in this article, the part I'd actually keep if you cut the rest is the restoration column and the pairing rule, and the reason is that it's the only part of the standard grief-journal advice that changes what you do rather than what you feel while doing it. The loss-oriented prompts are already everywhere. They're fine. They were never the gap.

And I'd hold the whole thing loosely. The pooled effect for this entire category of writing is small, the one properly randomised test in bereaved people came back null, and anybody telling you a notebook resolves a death is describing a product rather than a process. Write because you want the record, or because it's two in the morning and there's nothing else open. Both are sufficient reasons. Neither one is a clinical reason.

For material that goes darker than grief and into the parts of yourself the loss has uncovered, shadow work journaling covers the stop signals in more depth, and the broader guide to journal prompts explains why the phrasing of a question changes whether you end up analysing or looping.