diff --git a/proposals/4021-archive-controls.md b/proposals/4021-archive-controls.md
new file mode 100644
index 00000000000..ddb349564d0
--- /dev/null
+++ b/proposals/4021-archive-controls.md
@@ -0,0 +1,52 @@
+# MSC4021: Archive client controls
+
+The creation of archive.matrix.org indicates that search engine indexing of public Matrix rooms is a goal, but more
+granular control over how rooms should be indexed and displayed in search engine results should be granted to room
+admins.
+
+The current solution determines indexing eligibility based on `world_readable` `public` history visibility, but this is
+not an ideal solution because these settings only imply world readability within regular Matrix clients to most users,
+as opposed to the wider internet. Most alternative social media platforms provide separate settings for profile
+visibility and search engine indexing, for example.
+
+
+## Proposal
+
+Add an `m.room.archive_controls` state event where you can specify information about if and how you would like your
+room to be crawled. The [/publicRooms API](https://spec.matrix.org/v1.7/client-server-api/#get_matrixclientv3publicrooms)
+must relay this information to clients.
+
+| key | type | value | description | required
+|--|--|--|--|--
+| `archive` | boolean | | Whether the room should be included in room directory listings which are indended to be viewed by the public |
+| `robots` | [string] | Valid [robots meta rules](https://developers.google.com/search/docs/crawling-indexing/robots-meta-tag#directives) | A list of rules which should be included in a `robots` meta tag and/or [HTTP header](https://developers.google.com/search/docs/crawling-indexing/robots-meta-tag#xrobotstag-implementation) by public-facing clients. e.g. `["noarchive"]` or `["noindex", "nofollow"]`.
+| `via` | string | Hostname | A hostname which should be set as the canonical archive URL. e.g. `"archive.matrix.org"`.
+
+Public-facing clients like [matrix-public-archive](https://github.com/matrix-org/matrix-public-archive) should validate
+these rules before returning them in a response.
+
+When `archive` is `false`, clients which display a room directory intended for public internet consumption (e.g.
+matrix-public-archive or matrix-static) should exclude that room from being displayed. Clients which provide access
+to native Matrix users (e.g. Element) should ignore this setting.
+
+When `via` is specified, the client should return a [rel=canonical link element](https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls#rel-canonical-link-method)
+and/or a [rel=canonical HTTP header](https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls#rel-canonical-header-method)
+with the response pointing to the archive URL on the specified hostname. This prevents the Matrix.org public archive
+from returning duplicate content or taking precedence in search results over an organization's self-hosted archive.
+
+For example, if `via` is set to `"archive.example.net"` in `#main:example.net`, the page at
+https://archive.matrix.org/r/main:example.net/date/2023/05/28 should return this HTTP header:
+
+```
+Link: <https://archive.example.net/r/main:example.net/date/2023/05/28>; rel="canonical"
+```
+
+
+## Alternatives
+
+- [MSC2219](https://github.com/matrix-org/matrix-spec-proposals/pull/2291) could provide an alternative method of
+  specifying this information. However, this proposal includes the web archive metadata in the room directory API,
+  in order to access this information efficiently (this is a [requirement](https://github.com/matrix-org/matrix-public-archive/issues/47#issuecomment-1536938601)
+  for the matrix-public-archive project, for example). This proposal also allows rooms to opt-out of publicly accessible
+  room directories without clients like matrix-public-archive needing to join the room to read the state, and should
+  be interpreted by any client built for public web crawler access rather than [specific bots/clients](https://github.com/matrix-org/matrix-spec-proposals/pull/2291/files#diff-2b62d9e1c5ef21f7e10959da64da4000a69069b4dfb5d436db30d12c6bd23cb7R21-R23)