Most people shopping for an app for GTD have already read the book, tried three apps, and quietly gone back to a notes file. The method isn't the problem. Getting Things Done asks for five specific behaviours, and most task apps make two of them easy, one of them awkward, and two of them impossible.
So the useful question isn't which app is best. It's which of the five your app has to carry, and which you're willing to do in your head.
What an app for GTD has to do
David Allen's system runs on a loop: capture, clarify, organise, reflect, engage. An app can only really help with four of those, and it helps unevenly.
Capture has to be faster than the thought. If getting "call the dentist" into the system takes four taps and a category picker, you will not do it while walking to the car. You'll try to remember it, and that's the failure the whole method exists to prevent. A share sheet, a keyboard shortcut, a capture bar that's already focused — one of those has to exist.
Clarify has to be a separate step. This is the part apps get wrong most often. If the only way to add something is to file it immediately, then capture and clarify collapse into one action and capture slows down. You need a bucket that's allowed to be messy for a few hours.
Organise means a real hierarchy. A flat list of 200 items with tags on them is a filing cabinet with no drawers. GTD projects are anything needing more than one action, which for most people is forty or fifty live things. Those need to nest under something.
Reflect is the weekly review, and no app does it for you. What an app can do is make the review take twenty minutes instead of ninety, by keeping a someday shelf, a waiting-for filter, and a log of what actually got done.
Engage is just doing the work. The app's only job there is to not be in the way.
The five lists, and where they live in a task app
GTD's lists aren't features you turn on. They're jobs, and every app maps them somewhere — sometimes badly.
| GTD list | what belongs in it | where it lives in an app |
|---|---|---|
| In-tray | anything unprocessed, from "milk" to a half-formed idea | a real inbox, empty by default |
| Next actions | the single physical next step for each live project | today / next 7 days, or a saved filter |
| Projects | anything that takes more than one action | a project object with tasks inside it |
| Waiting for | delegated things you can't act on yet | a [waiting] tag plus a saved filter |
| Someday / maybe | not now, not never | a separate shelf, out of the daily view |
| Reference | the article, the quote, the PDF you'll need when you get there | usually nowhere, which is the whole problem |
That last row is where most GTD setups leak. The method says reference material belongs beside the project it serves. Task apps only store tasks, so the link ends up in a notes app, the PDF in Downloads, and the address in a text message. By the time you sit down to do the thing, you're hunting.
Tiago Forte's PARA is the same observation from the other direction: projects and areas hold both the actions and the material. If your app can't hold both, you're running two systems and reconciling them by hand.
Set up GTD in one sitting
You do not need a weekend. Ninety minutes gets you a working system; the rest is upkeep.
- Pick the app and stop shopping. Any app with an inbox, projects and a way to file things will work. Switching costs more than the difference between the two apps you're deciding between.
- Empty your head into the inbox. Everything — the passport, the birthday, the thing at work you keep half-thinking about. Don't sort while you do this. Twenty minutes, no editing.
- Create your areas first, not your projects. Work, home, money, health, whatever yours are. Four to six is normal. More than eight usually means some of them are projects.
- Turn every inbox item into either a project, a next action, a reference note, or nothing. "Nothing" is a legitimate answer for about a third of them.
- Write next actions as physical steps. "Passport" is not an action. "Find the old passport in the filing box" is. If you can't picture yourself doing it, it's a project, not an action.
- Move anything you can't start this month to the someday shelf. It stays in the system, it just stops shouting.
- Put a recurring weekly review on the calendar for Friday afternoon, and keep the first one to fifteen minutes. A short review you do beats a thorough one you skip.
Day two will feel worse than day one. The inbox refills, and you discover that half your projects had no real next action. That's the system working, not failing.
What a free GTD app costs you
Almost every free GTD app is free at the tier where GTD doesn't work yet. The common walls: a cap on projects, no saved filters, no recurring tasks, or reminders behind a subscription. Each one hits a different part of the loop, and the waiting-for filter is usually the first casualty.
Worth checking before you commit: can you export? A GTD system is three years of accumulated structure. If the export is a screenshot, you don't own it. And if the app can produce a CSV or JSON dump, you can hand that file to an AI assistant and have it rebuilt somewhere else in a few minutes — which quietly makes lock-in a much smaller risk than it used to be.
The weekly review is the part everyone drops
The review is the only step with no immediate payoff, so it's the one that goes first. Three things make it survivable.
Keep a real someday shelf, so the review is mostly moving things off the active list rather than onto it. Keep a log of completed work, so fifteen minutes of the review is reading evidence that the week happened. And do it at the same time every week, badly, rather than properly once a month.
If your list is honest, the review is short. If the review keeps taking an hour, that's usually a signal that next actions are being written as projects.
Where init.Tasks fits
init.Tasks is a to-do list that looks like a terminal, on iOS, macOS and web. Areas hold projects, projects hold to-dos and subprojects, and the inbox holds anything you haven't decided about yet. Quick add takes plain language — renew passport 7 oct #paperwork files itself.
Two things matter for GTD specifically. Start date and deadline are separate fields, so "could start Monday" doesn't pretend to be "due Monday" — that distinction is what keeps the next-actions list honest. And reference material lives next to the tasks, so the ryokan link and "book the ryokan" are in the same project.
An MCP server is coming in august: connect Claude or ChatGPT and it can read your projects and file things for you. Nothing about GTD requires that. It's mostly useful for the boring parts, like rebuilding a system you exported from somewhere else.
The beta waitlist is open and invites go out mid-to-late august. Everything is free while in beta. The feature list is the honest version of what exists today.
FAQ
What is the best app to use for GTD?
The one whose capture step you'll actually use. Every serious contender — Things, OmniFocus, Todoist, TickTick, init.Tasks — implements projects and lists competently. They differ on capture speed, whether reference material has a home, and how much the weekly review costs you. Pick on those three, not on feature counts.
What is a GTD app?
A task app whose structure matches the Getting Things Done method: an inbox for unprocessed capture, projects for anything multi-step, next actions written as physical steps, a waiting-for list, and a someday shelf. Nothing is GTD-specific in the code — it's about whether the app makes those five lists natural or awkward.
Is GTD still relevant?
The mechanics are, and they predate the software by decades. What has aged is the assumption that reference material lives in a filing cabinet. Most of what you need to finish a project today is a link, a screenshot or a message, and the method never said where those go. That gap is why so many GTD setups end up spread across three apps.
What is the best GTD app for Mac?
On the Mac the field narrows to apps with a real desktop client and a global capture shortcut — OmniFocus and Things are the long-standing answers. init.Tasks ships a native macOS app alongside iOS and web, with keyboard-first navigation and the same data on all three. It's in private beta rather than on sale, so treat it as a candidate rather than a recommendation for now.