How to run a weekly review

What a GTD weekly review is

A GTD weekly review is a scheduled pass over your whole system to bring it back in line with reality. You are not doing the work. You are deciding what is still true: what landed, what stalled, what no longer matters.

Copy this list into Today as a repeating task, or print it. Eleven steps, three phases.

Get clear

• Collect loose papers and notes
• Get every inbox to zero
• Empty your head onto the list

Get current

• Review next-action lists
• Review the previous calendar
• Review the upcoming calendar
• Review waiting-for
• Review project lists (every project has a next action)
• Review any other checklists you keep

Get creative

• Review someday/maybe
• Be creative and courageous: what would you add if nothing were in the way?

The official one-page PDF is the source. This page is the same eleven steps with timing, stall recovery, and how to keep the review short.

David Allen calls it the critical success factor for keeping a system you trust. A list you do not review becomes a list you do not believe, and a list you do not believe is one you stop opening.

For printable checklists and trigger lists that pair with this guide, use the free GTD weekly review resources.

The eleven steps, grouped

The official checklist is a free one-page PDF from the David Allen Company. Two details most write-ups get wrong: there are two separate calendar steps, backwards and forwards, and the third phase is Get Creative, which is the one people drop when they treat the review as tidying up.

Get clear: every inbox at zero

Empty every place things accumulate. Your task inbox, the physical pile, the notes app, the screenshots, the browser tabs left open as reminders.

Then empty your head: write down every commitment that surfaces without deciding what to do about it yet. This feels unproductive and does the most work.

Masicampo and Baumeister found unfinished goals cause intrusive thoughts during unrelated tasks, and that a specific plan removes the interference, but only for people whose plans were earnest enough that they acted on them. A vague note buys no quiet. Deciding the next step does.

The older Zeigarnik effect is quoted more confidently than it deserves. Zeigarnik found unfinished tasks about 90% more likely to be recalled, but Seifert and Patalano document “frequent replications and non-replications”, and their own first experiment found the reverse. The effect returned when subjects hit a genuine wall rather than being interrupted arbitrarily, which fits the review: the problems you got stuck on are the ones that keep surfacing.

Get current: every project needs a next action

Read your calendar backwards first. Last week’s meetings generate follow-ups you never captured, the most commonly lost commitments in any system. Then read forward far enough to catch what needs preparation.

Go through your waiting-for list and decide, item by item, whether to chase, wait, or drop. Then open every active project and confirm it has at least one next action. A project without one is stalled, and stalled projects stay invisible until a deadline makes them urgent.

Get creative: revisit what you parked

Read your someday list. Some items have quietly become relevant and should move to your active lists. Some are dead and should be deleted.

Deleting is the point. A someday list that only grows becomes a monument to guilt, at which point you stop reading it and the review loses its last phase.

Then add the ideas you have been half-having. This is the only part that is not maintenance.

How long, and how often

The David Allen Company suggests 1 to 2 hours for a thorough review. That figure comes from Kelly Forrister, a senior coach at the company, not from Allen in the checklist, and you will see it quoted as though it were.

The same post is more useful on cadence: every 7 to 10 days, not strictly weekly, and Friday morning as the most common choice, with the advice to pick a time when you are “brain sharp, not brain toast.” If you have been failing at Sunday evenings, that is the reason.

Ignore two numbers. The “Friday afternoon” version has no primary source, and the 2½ hours on gettingthingsdone.com is a paid audio product’s runtime.

Treat one to two hours as the size of a thorough review rather than the price of entry. A review of a system that is already close to current takes twenty minutes, and the long ones are long because the previous review was skipped. The figure people quote is the recovery cost, not the maintenance cost.

Why people stop

Two reasons, neither of them discipline.

The review has no appointment, so it competes with real commitments and loses. Put it in the system it reviews, on a repeating date, at a named time. Gollwitzer and Sheeran’s meta-analysis of 94 tests found that specifying when, where and how in advance has a medium-to-large effect on whether something happens. “Weekly review, Friday 9am” is that structure. “I should review more often” is not.

The other reason: the system needs assembling before it can be reviewed. If next actions live in a filter you rebuild and someday items are a label you have to remember, step one of your review is tool maintenance. That tax is what most abandoned setups die of.

How Done Bear helps

Done Bear has GTD’s lists built in as the navigation. It will not run the review or tell you what to drop. What it removes is the assembly step, because the lists the checklist asks for are the five views you already use: Inbox for “get in to zero”, Upcoming for the forward calendar read, Anytime for next actions, Someday for the get-creative pass, and Logbook for what got finished.

Emptying your head produces a burst of items, so capture matters here. It comes from web, desktop, iOS, a Raycast extension, the CLI, and an AI assistant via MCP, all into one Inbox. Over MCP you can also ask an assistant to compare your Logbook against what is still in Anytime, which tells you what you have been avoiding.

No study evaluates the GTD weekly review as an intervention. The citations here support the mechanisms it plausibly exploits, not GTD itself, and anyone calling the review “scientifically proven” is overreaching.

When you might not need this: under twenty commitments that fit in your head, a weekly review is ceremony.

The shape is the same in other apps. Things 3 has these lists too, and the Things 3 comparison covers where they differ.

Start smaller than feels right

Three phases, eleven steps, one recurring appointment. If that is a lot, run only get clear and get current for the first month. A short review you do beats a thorough one you keep rescheduling.