diff --git a/content/en/docs/tasks/administer-cluster/highly-available-master.md b/content/en/docs/tasks/administer-cluster/highly-available-master.md
index 6e9bd16f680e2..c3a86e16752d8 100644
--- a/content/en/docs/tasks/administer-cluster/highly-available-master.md
+++ b/content/en/docs/tasks/administer-cluster/highly-available-master.md
@@ -31,7 +31,7 @@ To create a new HA-compatible cluster, you must set the following flags in your
 * `MULTIZONE=true` - to prevent removal of master replicas kubelets from zones different than server's default zone.
 Required if you want to run master replicas in different zones, which is recommended.
 
-* `ENABLE_ETCD_QUORUM_READS=true` - to ensure that reads from all API servers will return most up-to-date data.
+* `ENABLE_ETCD_QUORUM_READ=true` - to ensure that reads from all API servers will return most up-to-date data.
 If true, reads will be directed to leader etcd replica.
 Setting this value to true is optional: reads will be more reliable but will also be slower.