reading the database…
Drawn to living in community: ecovillages, cohousing, communes, land projects? Answer a few questions and we'll point you toward places worth exploring. You still do the real finding; this just gives you a first direction.
A tool for finding your way toward community. Every result shows you why it surfaced; the rest is yours to walk. Go visit, ask, stay a while.
Here on behalf of a community rather than looking for one? Add it to the map, it takes a minute.
✶ Want to be connected to more like this? We're weaving a living web of communities. Drop your email and we'll send the occasional note as it grows.
If you run or belong to a community that isn't here, this is the place to say so. Everything except the name and roughly where you are is optional. A person reads every submission, checks it against what's public, and writes back. Being listed is free, and you can ask to be removed at any time.
Thank you for adding to the map. It gets truer every time someone does.
Community Compass is a free starting map for people drawn to living in community: ecovillages, cohousing, communes, land projects, shared houses. It points a direction; it doesn't tell you where you belong. That part is yours to walk. This page explains how it works, where the data comes from, and where it's still thin, so you can trust it the right amount, and no more.
Built by Gustaf Palm, part of the regen-community body of work and the Animate Intelligence project, under the hyphae brand. It grew from a stubborn problem: the world is full of intentional communities, and almost no one drawn to them knows where to begin. The directories are scattered (one 2022 roundup catalogues 21 separate ones), and a lot of what they list no longer leads anywhere live. When we examined a 16,390-record index one by one, only 4,979 could be confirmed as currently active from public web evidence; 1,267 more are carried on a federation's own listing and labelled as such. Those 6,213 — the same set the Living Atlas maps — are what this tool searches. This is one doorway in: a place to start the search, not a service that finishes it for you.
The communities are drawn from the regen-community database: 16,390 entries gathered from public directories, federations and networks, then checked one by one. As of June 2026 every entry has been individually web-verified. Nothing here is a scraped list we never looked at. Contact details come from those same public directories and are obfuscated on the page; the plaintext only appears after you click to reveal it, so listings can't be harvested by bots.
The Compass only shows the gold set: 6,213 communities we've confirmed are active. Communities that came back rejected, dead-linked, unconfirmed or defunct are excluded entirely: a smaller map you can trust over a bigger one padded with ghosts. The networks and platforms we point to alongside them are held to a lighter bar: they're leads for going deeper, listed as-is rather than verified one by one. Coverage isn't uniform, and it tracks the directories we ingested rather than where communities actually are. The two largest country blocks are France (813) and the United States (796), then the United Kingdom (292), Canada (291), Russia (288) and Israel (285); Germany has about 90. Roughly a hundred of the French entries are Emmaüs solidarity communities. Read this as a map of our sources, not a map of the world. Communities we've verified but don't yet have coordinates for still appear in your results, marked precise location pending; they just don't show on the map until placed.
You answer nine short questions; the Compass surfaces communities that echo what you asked for, sorted into bands: Strong fit, Good fit, Partial fit. Every card shows its receipts: the specific reasons it surfaced, and where we're guessing (a hedged "·") versus knowing (a "✓"). Read the band for exactly what it is: how closely a place matches what you told us you wanted, not a rating of how good it is. A "Partial fit" isn't a lesser place; it answered fewer of your particular asks. We don't rank communities against each other.
Some honest gaps, because you deserve to know the state of the field you're entering. Nobody has ever measured a survival rate for modern communities; if someone quotes you "90% fail," that number is one practitioner's estimate from a 2003 book, about forming groups that never got started, not a study. The closest thing to a real measurement is narrower and says something different: across 26 archived snapshots of one cohousing directory (US-based, 1994 to 2017), 69% of the forming groups that resolved either way stopped before anything was built; count the groups still forming and it's 55%. That measures listings going quiet in a self-maintained directory, not communities dissolving, and the compiler's own caveat is that a stopped listing can simply mean a name change. Nobody has put the two sides together. The places have been counted, partially: the Foundation for Intentional Community lists 1,000+ worldwide, the Global Ecovillage Network claims it reaches around 6,000 in 137 countries, our own map holds 4,979 web-verified. The seekers have only ever been sampled, once nationally (a 2016 US survey of 1,000 people), and project by project (400-plus people came forward for 11 homes at one London cohousing scheme). No one has counted seekers against places across the field. And no matching tool anywhere, including this one, has proven with data that its matches beat a good conversation and a visit. That's why the Compass gives you bands and reasons instead of a compatibility score: a score would be dressing a guess up as a measurement. The one time researchers really tried — machine learning over 100+ things speed-daters said about themselves before meeting (US students, 2005 and 2007 data) — it predicted nothing about who would want whom; what did predict it was meeting. What the little evidence there is does say: fit is real. At one cohousing community tracked for a decade, a third of the founding residents remained after ten years, and "not a good fit, too much work" was among the reasons people left. Leaving is normal. The match is step one; the visit is the test.
If you run or belong to a community that isn't listed here, tell us about it. A person reads every submission and writes back. Listing is free and removal on request is always honoured.
If you run or belong to a community listed here and something's wrong (out of date, mis-described, wrongly placed, or you'd like to be removed), reach out and we'll fix it. This map gets more accurate every time a real person tells us what we got wrong.
A starting map. The walking is yours.
Plain language, no legalese. The short version: we collect as little as possible, we don't sell anything, and you can have your data corrected or removed by asking.
The only personal data we collect from seekers is an email address, and only if you choose to enter one. It's sent to our form service (FormSubmit) and reaches our inbox so we can send the occasional note as the map grows. We don't sell or share it, and you can leave anytime by replying "remove."
Contact details for the listed communities come from public directories and networks. They're obfuscated on the page; the plaintext only appears after a human clicks to reveal it, so bots can't harvest them. Some communities are in Europe, so we treat this data with GDPR in mind.
If your community is listed and you'd like it corrected or removed, or you want to know what we hold, email gustaf@islands-of-coherence.com and we'll act on it. No account, no forms to fill.
Least data, held lightly.