diff --git a/designs/priorityqueue.md b/designs/priorityqueue.md
new file mode 100644
index 0000000000..ef1f7588a6
--- /dev/null
+++ b/designs/priorityqueue.md
@@ -0,0 +1,110 @@
+Priority Queue
+===================
+
+This document describes the motivation behind implementing a priority queue
+in controller-runtime and its design details.
+
+## Motivation
+
+1. Controllers reconcile all objects during startup to account for changes in
+   the reconciliation logic. Some controllers also periodically re-reconcile
+   everything to account for out of band changes they do not get notified for,
+   this is for example common for controllers managing cloud resources. In both
+   these cases, the reconciliation of new or changed objects gets delayed,
+   resulting in poor user experience. [Example][0]
+2. There may be application-specific reason why some events are more important
+   than others, [Example][1]
+
+## Proposed changes
+
+Implement a priority queue in controller-runtime that exposes the following
+interface:
+
+```go
+type PriorityQueue[T comparable] interface {
+    // AddWithOpts adds one or more items to the workqueue. Items
+    // in the workqueue are de-duplicated, so there will only ever
+    // be one entry for a given key.
+    // Adding an item that is already there may update its wait
+    // period to the lowest of existing and new wait period or
+    // its priority to the highest of existing and new priority.
+    AddWithOpts(o AddOpts, items ...T)
+
+    // GetWithPriority returns an item and its priority. It allows
+    // a controller to re-use the priority if it enqueues an item
+    // again.
+    GetWithPriority() (item T, priority int, shutdown bool)
+
+    // workqueue.TypedRateLimitingInterface is kept for backwards
+    // compatibility.
+    workqueue.TypedRateLimitingInterface[T]
+}
+
+type AddOpts struct {
+    // After is a duration after which the object will be available for
+    // reconciliation. If the object is already in the workqueue, the
+    // lowest of existing and new After period will be used.
+    After time.Duration
+
+    // Ratelimited specifies if the ratelimiter should be used to
+    // determine a wait period. If the object is already in the
+    // workqueue, the lowest of existing and new wait period will be
+    // used.
+    RateLimited bool
+
+    // Priority specifies the priority of the object. Objects with higher
+    // priority are returned before objects with lower priority. If the
+    // object is already in the workqueue, the priority will be updated
+    // to the highest of existing and new priority.
+    //
+    // The default value is 0.
+    Priority int
+}
+```
+
+In order to fix the issue described in point one of the motivation section,
+we have to be able to differentiate events stemming from the initial list
+during startup and from resyncs from other events. For events from the initial
+list, the informer emits a `Create` event whereas for `Resync` it emits an `Update`
+event. The suggestion is to use a heuristic for `Create` events, if the object
+in there is older than one minute, it is assumed to be from the initial `List`.
+For the `Resync`, we simply check if the `ResourceVersion` is unchanged.
+In both these cases, we will lower the priority to `LowPriority`/`-100`.
+This gives some room for use-cases where people want to use a priority that
+is lower than default (`0`) but higher than what we use in the wrapper.
+
+```go
+// WithLowPriorityWhenUnchanged wraps an existing handler and will
+// reduce the priority of events stemming from the initial listwatch
+// or cache resyncs to LowPriority.
+func WithLowPriorityWhenUnchanged[object client.Object, request comparable](u TypedEventHandler[object, request]) TypedEventHandler[object, request]{
+}
+```
+
+```go
+// LowPriority is the priority set by WithLowPriorityWhenUnchanged
+const LowPriority = -100
+```
+
+The issue described in point two of the motivation section ("application-specific
+reasons to prioritize some events") will always require implementation of a custom
+handler or eventsource in order to inject the appropriate priority.
+
+## Implementation stages
+
+In order to safely roll this out to all controller-runtime users, it is suggested to
+divide the implementation into two stages: Initially, we will add the priority queue
+but mark it as experimental and all usage of it requires explicit opt-in by setting
+a boolean on the manager or configuring `NewQueue` in a controllers opts. There will
+be no breaking changes required for this, but sources or handlers that want to make
+use of the new queue will have to use type assertions.
+
+After we've gained some confidence that the implementation is useful and correct, we
+will make it the default. Doing so entails breaking the `source.Source` and the
+`handler.Handler` interfaces as well as the `controller.Options` struct to refer to
+the new workqueue interface. We will wait at least one minor release after introducing
+the `PriorityQueue` before doing this.
+
+
+* [0]: https://youtu.be/AYNaaXlV8LQ?si=i2Pfo7Ske6rTrPLS
+* [1]: https://github.com/cilium/cilium/blob/a17d6945b29c177209af3d985bd82cce49eed4a1/operator/pkg/ciliumendpointslice/controller.go#L73