You have four real options for a finished journal: keep it unread, read it once and then keep it, pull out the five lines worth keeping and store the rest, or destroy it. All four are defensible, and which one is right depends on what kind of writing it was. Journals kept to process difficult things are often better destroyed or left unread; journals kept to record a life are worth reading and keeping. If you write in an app rather than a notebook, the January question changes shape entirely, and that is covered below too.
Why does this question come up in January?
Because a notebook finishing and a year finishing usually land near each other, and because January is when people decide to start fresh. The new notebook is bought, and the old one is suddenly an object with no clear job.
There is also a less comfortable reason. A lot of people have several years of journals they have never reread, are slightly afraid to reread, and cannot bring themselves to throw away. The pile grows by one every January and the decision gets postponed every January.
Making the decision once, deliberately, is worth more than making it badly eleven times.
What are the four options?
Pick by what the journal was for, not by what feels most virtuous. Most people have written more than one kind over the years, and they do not all deserve the same treatment.
| What the journal was | Best option | Why |
|---|---|---|
| A record of days, people and events | Read once, then keep | This is the kind that becomes genuinely valuable in ten years |
| Processing something hard | Keep unread, or destroy | Rereading a difficult stretch line by line is rarely useful and often makes it worse |
| Venting, mostly complaints | Pull out anything factual, then destroy | It served its purpose the day it was written |
| Planning, goals, work notes | Extract the decisions, then store | The useful part is a page, not the notebook |
| Someone else's private information in it | Decide with that in mind | Other people did not consent to being on record |
How do you read a year back without losing an evening?
Read it in one sitting, with a notebook beside you, and do not read every word. Skim, and stop only where something makes you pause.
Write down four things as you go, nothing else.
- Facts you had forgotten. Dates, names, what actually happened when. This is the part memory loses first and the part that is worth most later.
- Anything that repeats. A worry that appears in March, July and November is telling you something a single entry cannot.
- Decisions you made, and what you thought at the time. In a year you will remember the decision and not the reasoning.
- Five lines worth keeping. Genuinely five. If you find thirty, you are copying rather than selecting.
- Nothing about how the writing reads. Cringing at your own sentences from March is a reliable way to stop the exercise early.
When is destroying a journal the right call?
More often than people expect, and it is worth saying so plainly, because the usual advice treats keeping everything as automatically better.
Destroy it when the writing was a way of getting through something and keeping it means keeping the thing accessible. Destroy it when it contains someone else's private information that they would not want recorded. Destroy it when the thought of anyone finding it is doing more harm than the record is doing good, which is a real and reasonable concern rather than paranoia.
The middle path is worth knowing: copy out the few lines that matter, the facts and the dates, then destroy the rest. You keep the record and lose the raw material.
What is not a good reason is embarrassment at who you were. That version of you is the point of the archive.
Should you start a new journal in January?
With a paper notebook, the calendar decides this for you eventually, and January is as good a moment as any. Finishing on the thirty-first and starting on the first is neat, and neatness is genuinely useful when you come to find something in five years.
With an app, the question is different and mostly imaginary. There is no notebook to fill, so starting over in January means either deleting what came before or creating a second account, and both of those cost you the only thing a digital journal has over paper: the ability to search across years.
The useful version of a fresh start in an app is not a new journal. It is a reintroduction entry: who is around now, what you are in the middle of, what you want to be different. That gives you a clean marker in the timeline without cutting it.
How should you store the ones you keep?
For paper: one box, dated on the spine, somewhere dry and not visible. Boxed and labeled beats a shelf, because a shelf invites the pile to grow without anyone deciding anything.
For digital: export once a year and keep the file with your other important documents. An export you have never opened is a guess rather than a backup, so open it and read three entries from three different months before you file it.
The failure mode for digital journals is not losing the file, it is losing access: an account you cannot get into, a subscription that lapsed, an app that shut down. A plain text or Markdown export sitting in a folder you control avoids all three.
What is the point of keeping any of it?
One concrete thing, mostly: it is the only accurate record you have of what you thought at the time. Memory updates itself constantly and quietly, rewriting past opinions to match current ones, so a journal is the only place your earlier reasoning survives intact.
The second reason is pattern. A single year is a story; five years is evidence. The same worry appearing every autumn, the same kind of argument with the same kind of person, the same decision made three times, are all invisible inside one notebook and obvious across several.
Reflectize is an AI journal designed around going back: it holds on to the people, plans and moments in your entries and brings them up when they matter again. Available on iPhone and Android. The difference it makes to this particular question is that you do not need an annual reading session to find the patterns, because the searching and the surfacing is the app's job rather than yours.
What if you have ten years of them and have read none?
Do not start at the beginning. Pick one year, ideally one you have a reason to be curious about, and read only that. A full archive read as a project is a plan that stalls in week two and leaves the pile exactly where it was.
After one year, decide the rule for the rest. Most people find that one of two things is true: either the reading was worth it and the rest can be done slowly, one year each January, or it was not, in which case the honest move is to keep them unread as an archive rather than as a task.
Either answer is fine. What costs you something is the third state, where they are neither read nor decided about, and the pile quietly means an obligation every time you see it.
This piece is part of self-awareness and patterns. Single entries record a day. Read together, they show a shape. These pieces are about looking back across weeks and months and noticing what keeps coming up.