How to Journal Between Therapy Sessions So You Bring Something Back

Most people in therapy who journal arrive at their session with a week of writing and nothing they can use. I see it constantly across Rosebud's users. The problem is not the writing. It is that the writing was never the deliverable.

Here is what a week of journaling is supposed to produce: three things. One recurring theme. One moment you could not place. One thing you are avoiding. Each one points back at an entry you have not edited. Everything else stays in the journal, which is where it belongs.

I'm Chrys Bader, CEO of Rosebud. I've spent over a decade building consumer products - from Google to Secret to now, an AI journal that's helped 100,000+ people process their thoughts.

83% of Rosebud subscribers are or were in therapy. I am not writing this for people looking for an alternative to the room. I am writing it for people who already have a therapist, already value the hour, and want the six days in between to actually feed it.

Which means the week between your sessions has a job: produce material worth selecting from, and then select.

What a therapy journal is for

A therapy journal is the writing you do between sessions to catch what the session started and carry that material back into the room. That is the whole job. A therapy journal is not a diary and it is not a transcript. If your entries are not producing something you can use in the next 50 minutes, the journal is running as a hobby that happens to be about therapy.

The hour moves fast. You leave with more than you can hold, and memory quietly edits the rest - usually in the direction that makes you look reasonable. Writing catches the material before that editing happens, which is the whole therapeutic mechanism.

None of this depends on your tool, and I have written about journaling apps for adults in therapy elsewhere.

The first 24 hours are for capturing while you are still full

Write within a day of the session, while you still have the material. Not a summary. Whatever is loudest - the thing you said that surprised you, the thing you did not say, the sentence of your therapist's that will not leave. Five honest minutes beats twenty polished ones. You are catching raw material for later, not processing it now.

1. Write within 24 hours of the session, before the day ends.
2. Start with whatever is loudest, not with what happened first.
3. Stop at five minutes if five minutes is what you have.
4. Leave it unedited, in the words and feelings you actually had.

The reason for the speed is unglamorous. You are still emotionally full when you walk out, and the feelings have not sorted themselves into a story yet. A day later you have a story. Two days later, a summary of it.

One caveat, not a footnote. Baikie and Wilhelm's review of expressive writing reports it was detrimental for adult survivors of childhood abuse and for a small sample of eight Vietnam veterans with PTSD. Eight people, and I will not inflate that into a population claim. But if writing straight after a session reliably leaves you worse for days rather than hours, that is information for the room. Raise it with your therapist rather than pushing through alone.

Days two to four are the flat middle where the practice dies

Days two to four are where journaling between sessions actually fails. The session has faded and the next one is not close enough to create urgency, so there is no pressure in either direction. The fix is a smaller ask, not more discipline - one line a day, enough to keep the week's material catchable. Deep writing is for the days something lands hard, not for every day.

Here is why the flat middle is load-bearing rather than filler: material you never caught on Wednesday is material you cannot select from on Sunday. Miss it and you arrive at the read-back with two good entries and a hole.

The famous protocol was never built for this. Baikie and Wilhelm describe Pennebaker and Beall's original paradigm as 15 minutes of writing on four consecutive days about a single stressful event. Four days. One event. The standard research paradigm since then runs three to five sessions of 15 to 20 minutes. Week 40 of an open-ended weekly practice was never the question any of it answered.

So stop treating a four-day laboratory protocol as a daily quota you are failing. Write deeply when something lands. Write one line when nothing does. The streak was never the point. The material is.

How to tell reflection from rumination

Start by asking where the writing is going. Reflection turns inward to solve something. Brooding compares your situation against a standard you have not met, and Treynor and colleagues found brooding predicts more depression a year later. Run this test on your own entries. Is this going somewhere new, or is it draft seven of the same paragraph?

Reflection, in Treynor and colleagues' work, is a purposeful turning inward to engage in cognitive problem solving to alleviate one's depressive symptoms. On the page, reflection looks like a sentence you had not written before.

Brooding is a passive comparison of one's current situation with some unachieved standard. On the page, brooding looks like the same paragraph again, with better adjectives.

Both come with limits. The finding is correlational and psychometric - a trait measure, not a study of journaling. It does not show journaling causes brooding, or that any technique flips brooding into reflection. The brooding result is the firm one. The authors are more cautious about reflection's benefit, and so am I.

So test your entries rather than diagnose yourself. Read the last three. If each gets somewhere the previous one did not, emotional processing is happening. If all three circle the same unmet standard, flag it for the room - it says nothing about what is wrong with you.

Elle, a Rosebud subscriber of seven months, put it better than any brain scan I have read: "There's like a gap between whatever I experienced, the feeling, and then my brain taking it, spinning it, and creating a story around it. There's like this three-second window, and I hadn't had that like pause before."

That pause is what you are testing for. Elle is describing what it felt like, not proof of anything.

The last 24 hours are for reading back, not writing more

In the day before your session, read the week instead of adding to it. Read, mark, and stop. Baikie and Wilhelm report that expressive writing usually produces a short-term increase in distress, which is exactly what you do not want the night before you walk in. The writing is already done. The last 24 hours are for selection.

1. Read the week in one sitting, the day before.
2. Mark anything that appears more than once.
3. Mark anything you skipped past quickly.
4. Stop. Do not tidy, expand, or rewrite what you find.

Reading back is not risk-free either. It is lower-risk than deep writing the night before, which is the actual claim - and the only way to see the patterns no single entry contains.

Here is where the tool changes the work, and I will be straight about it. Rosebud produces weekly reports across your entries, surfacing patterns and progress over a week rather than a single sitting. Kyle, a subscriber of two years, sends his to his wife and his therapist. Note what he sends: the report, not the entries.

You can run the same pass with a paper notebook and a highlighter. It takes longer and you will miss more, which is a real trade, not a pitch. Howes notes patients are sometimes amazed at how far they have come, and sometimes surprised to find they are barking up the same tree.

What to bring back from a week of journaling

Bring three things to your next session, not a notebook. One recurring theme. One moment you could not place. One thing you are avoiding. Each item is a pointer, not a summary - you select the entry and leave the raw words exactly as written. Everything that does not fit those three slots stays in the journal.

• One recurring theme. The thing that showed up on Tuesday and again on Friday without you planning it.
• One moment you could not place. A reaction that was louder than the event that caused it.
• One thing you are avoiding. The entry you started and did not finish.

That is three sentences of preparation, and it is deliberately smaller than what you wrote. Selection is the work. Polishing is the thing that looks like the work.

If two candidates compete for a slot, take the one you would rather not raise - and take it into the room, not further into the journal. You are flagging it for your therapist, not processing it alone on a Sunday night. That distinction is the whole safety margin.

That also settles the sharing question. You do not hand over the notebook. You hand over three pointers into it.

None of this is new, and I would rather name it than brand it. What you are doing is session agenda-setting, standard practice in CBT for decades. The only thing I am adding is where the agenda comes from: your own unedited week, selected down to three items.

What happens once you are seated - which of the three to lead with, how hard to push - is your therapist's ground, not mine.

Do not arrive with a briefing document

The failure mode here is not laziness. It is diligence. A typed chronological account of your week, cleaned up and organized, is the wrong artifact. Ryan Howes puts it plainly in Psychology Today - write as if no one else will ever read it, because writing for an audience risks getting lost in the performance. A polished account is already a performance. Bring the mess with a map, not a manuscript.

The reader who does this is not lazy. You did the wrong work diligently, which is why the habit is hard to spot. It feels like preparation.

Howes gives the constraint on sharing too. Bring it to session if you like, he says, but for highlights, not as a script - he does not want to do therapy with your journal. A script is a document you can perform. Three pointers into unedited entries is one you cannot, which is exactly why it works.

Match the method to the therapy you are actually in

The right journaling method depends on the therapy you are in - and that includes everything I have just told you. A prepared agenda is a well-functioning CBT session. In psychodynamic work, arriving with an agenda can be the defense itself. The floor rule settles it. If your therapist assigned something, that assignment wins, and free journaling only catches what the form has no box for.

That floor rule is not mine to override. It is the therapeutic contract you already have. Journal around it, not instead of it.

Howes is blunt about the edge case: he would not recommend journaling for someone in analysis. If you are in that work and your therapist has never raised it, that silence is more likely a clinical position than an oversight. Ask them. And if DBT diary cards or an EMDR protocol already set your agenda, that question is answered.

What the evidence actually supports

The honest numbers are smaller than the ones usually quoted. Sohal and colleagues pooled 20 randomized trials. The 9% anxiety reduction that circulates from that paper carries a confidence interval running from -0.19 to 0.01. The interval crosses zero, so the result is not statistically significant. The same authors still recommend journaling as an adjunct to treatment. Both of those are true at once, and neither one survives without the other.

Here is the rest of that 2022 meta-analysis of journaling for mental health, since you should have it rather than a number I picked. Across 31 outcomes, Sohal and colleagues found PTSD symptoms down 6%, and that one is statistically significant. Depression came in at 4%, with a confidence interval touching zero. Heterogeneity ran above 70%, so the trials were not measuring the same thing the same way.

Their conclusion is the part worth carrying. Given "the low risk of adverse effects, low resource requirement and emphasis on self-efficacy," they write, "primary care providers should consider this as an adjunct therapy to complement current evidence-based management." Adjunct. Complement. That is the peer-reviewed version of the position I have been arguing the whole way through.

So the mental health research supports the direction, not the dose. The cadence I have just given you is built from that plus what I have watched work. It has not been tested as a protocol. I am not going to tell you otherwise.

When writing to something that answers back helps

Here is the uncomfortable part for me. Rosebud is a conversation, and a conversation has an audience - which is the exact thing Howes warns against. When racing thoughts need a shape, a responsive prompt helps. When you are already performing for your therapist, a second audience makes the performance easier, not harder. Paige, a Rosebud subscriber with a psychology degree, wanted something that could not be charmed by her.

Her words: "I'm really good at tricking therapists... I want something that can't be charmed by me."

The honest answer depends on which failure mode you are already in. If you freeze at a blank page, something that answers back gets you writing. If you are already fluent, already good at the well-formed account, a responsive interlocutor is one more thing to be fluent at.

I am not going to resolve that for you. Test it the way you test everything else: read your entries back and ask whether they sound like thinking or performance. If you want the comparison, I have written up the apps to use between therapy sessions separately.

Start with the next 24 hours

Do not adopt a system. After your next session, before the day ends, write down the loudest thing - five minutes, unedited. That is the whole first step. The cadence and the read-back only matter once there is material to read back, and material only exists if the first 24 hours happened. Everything else is downstream of that.

Put it wherever you will actually put it. A notes app, the back of a receipt, a paper notebook from 2019. I built Rosebud partly because the hour after a session is when typing is the last thing you want to do, so if talking is easier, use that. The capture is the point. The tool is optional.

Your next session is already on the calendar. The week in front of it is the part you get to decide about.