diff --git a/CHANGELOG.next.asciidoc b/CHANGELOG.next.asciidoc
index db784002f81..815b8d809f1 100644
--- a/CHANGELOG.next.asciidoc
+++ b/CHANGELOG.next.asciidoc
@@ -385,6 +385,7 @@ https://github.com/elastic/beats/compare/v7.0.0-alpha2...master[Check the HEAD d
 - [Kubernetes] Remove redundant dockersock volume mount {pull}22009[22009]
 - Revert change to report `process.memory.rss` as `process.memory.wss` on Windows. {pull}22055[22055]
 - Add a switch to the driver definition on SQL module to use pretty names {pull}17378[17378]
+- Add interval information to `monitor` metricset in azure. {pull}22152[22152]
 
 *Packetbeat*
 
diff --git a/metricbeat/docs/modules/azure.asciidoc b/metricbeat/docs/modules/azure.asciidoc
index 8ec43e4be43..08a14132e8f 100644
--- a/metricbeat/docs/modules/azure.asciidoc
+++ b/metricbeat/docs/modules/azure.asciidoc
@@ -97,7 +97,7 @@ https://login.microsoftonline.de for azure USGovernmentCloud
 [float]
 === `monitor`
 This metricset allows users to retrieve metrics from specified resources. Added filters can apply here as the interval of retrieving these metrics, metric names,
-aggregation list, namespaces and metric dimensions.
+aggregation list, namespaces and metric dimensions. The monitor metrics will have a minimum timegrain of 5 minutes, so the `period` for `monitor` metricset should be `300s` or multiples of `300s`.
 
 [float]
 === `compute_vm`
diff --git a/x-pack/metricbeat/module/azure/_meta/docs.asciidoc b/x-pack/metricbeat/module/azure/_meta/docs.asciidoc
index 9065f2f283f..78e932a0153 100644
--- a/x-pack/metricbeat/module/azure/_meta/docs.asciidoc
+++ b/x-pack/metricbeat/module/azure/_meta/docs.asciidoc
@@ -89,7 +89,7 @@ https://login.microsoftonline.de for azure USGovernmentCloud
 [float]
 === `monitor`
 This metricset allows users to retrieve metrics from specified resources. Added filters can apply here as the interval of retrieving these metrics, metric names,
-aggregation list, namespaces and metric dimensions.
+aggregation list, namespaces and metric dimensions. The monitor metrics will have a minimum timegrain of 5 minutes, so the `period` for `monitor` metricset should be `300s` or multiples of `300s`.
 
 [float]
 === `compute_vm`