diff --git a/FAQ.md b/FAQ.md
new file mode 100644
index 0000000..5015feb
--- /dev/null
+++ b/FAQ.md
@@ -0,0 +1,37 @@
+List of questions:
+
+[[_TOC_]]
+
+# Will there be an iOS version of Briar?
+
+Short answer: Probably not.
+
+Long answer: [We're looking into whether an iOS version is feasible](https://code.briarproject.org/briar/briar/issues/445).
+Briar needs to run in the background to receive messages from contacts, and iOS has
+much tighter **restrictions on background apps** than Android (though
+Android's getting stricter).
+
+A typical iOS messaging app would use a push notification to wake the
+app when a message is received, but this exposes metadata to Apple's
+push notification service and the app developer's push gateway.
+
+To reduce the metadata leakage, the sender of the message could connect
+to the push gateway via Tor - but the push gateway and APNS would still
+know who was receiving the notification. In a publish-subscribe network
+like Briar, if a bunch of people always receive notifications whenever
+the Revolutionary Planning Council updates its blog, then even if we
+don't know who sent the notifications, we've learned something sensitive
+about the recipients.
+
+If we don't use push notifications then the best Apple allows us to do
+is wake up every 15 minutes and check for messages. But maybe the sender
+won't be online when we check (their 15 minute intervals might not be
+aligned with ours - clocks aren't perfect). So we need somewhere for the
+sender to store the message until our next check.
+
+We're working on a piece of software called a mailbox that will receive
+encrypted messages over Tor and store them until the owner collects
+them. It's designed to run on a spare Android device, laptop, or
+Raspberry Pi that's plugged into power and internet. Once that's ready
+it will make an iOS app more feasible. But we are not sure if the 15 minute
+delay will be a deal-breaker.
\ No newline at end of file