diff --git a/src/sdam/monitor.rs b/src/sdam/monitor.rs
index 9a530cf1d..3a1627a57 100644
--- a/src/sdam/monitor.rs
+++ b/src/sdam/monitor.rs
@@ -275,7 +275,11 @@ impl Monitor {
         let start = Instant::now();
         let result = tokio::select! {
             result = execute_hello => match result {
-                Ok(reply) => HelloResult::Ok(reply),
+                Ok(mut reply) => {
+                    // Do not propagate server reported cluster time for monitoring hello responses.
+                    reply.cluster_time = None;
+                    HelloResult::Ok(reply)
+                },
                 Err(e) => HelloResult::Err(e)
             },
             r = self.request_receiver.wait_for_cancellation() => {
diff --git a/src/test/spec/sessions.rs b/src/test/spec/sessions.rs
index 798d13a99..5e2c8d3ed 100644
--- a/src/test/spec/sessions.rs
+++ b/src/test/spec/sessions.rs
@@ -4,6 +4,7 @@ mod sessions_not_supported_skip_local; // requires mongocryptd
 use std::{
     future::IntoFuture,
     sync::{Arc, Mutex},
+    time::Duration,
 };
 
 use futures::TryStreamExt;
@@ -12,12 +13,18 @@ use futures_util::{future::try_join_all, FutureExt};
 use crate::{
     bson::{doc, Document},
     error::{ErrorKind, Result},
-    event::command::{CommandEvent, CommandStartedEvent},
+    event::{
+        command::{CommandEvent, CommandStartedEvent},
+        sdam::SdamEvent,
+    },
     test::{
         get_client_options,
+        log_uncaptured,
         server_version_gte,
         spec::unified_runner::run_unified_tests,
+        topology_is_load_balanced,
         topology_is_sharded,
+        Event,
     },
     Client,
 };
@@ -197,3 +204,84 @@ async fn implicit_session_after_connection() {
 }
 
 // Prose tests 18 and 19 in sessions_not_supported_skip_local module
+
+// Sessions prose test 20
+#[tokio::test]
+async fn no_cluster_time_in_sdam() {
+    if topology_is_load_balanced().await {
+        log_uncaptured("Skipping no_cluster_time_in_sdam: load-balanced topology");
+        return;
+    }
+    let mut options = get_client_options().await.clone();
+    options.direct_connection = Some(true);
+    options.hosts.drain(1..);
+    let heartbeat_freq = Duration::from_millis(10);
+    options.heartbeat_freq = Some(heartbeat_freq);
+    let c1 = Client::for_test()
+        .options(options)
+        .min_heartbeat_freq(heartbeat_freq)
+        .monitor_events()
+        .await;
+
+    // Send a ping on c1
+    let cluster_time = c1
+        .database("admin")
+        .run_command(doc! { "ping": 1 })
+        .await
+        .unwrap()
+        .get("$clusterTime")
+        .cloned();
+
+    // Send a write on c2
+    let c2 = Client::for_test().await;
+    c2.database("test")
+        .collection::<Document>("test")
+        .insert_one(doc! {"advance": "$clusterTime"})
+        .await
+        .unwrap();
+
+    // Wait for the next (heartbeat started, heartbeat succeeded) event pair on c1
+    let mut events = c1.events.stream();
+    const TIMEOUT: Duration = Duration::from_secs(1);
+    crate::runtime::timeout(TIMEOUT, async {
+        loop {
+            // Find a started event...
+            let _started = events
+                .next_match(TIMEOUT, |ev| {
+                    matches!(ev, Event::Sdam(SdamEvent::ServerHeartbeatStarted(_)))
+                })
+                .await
+                .unwrap();
+            // ... and the next heartbeat event after that ...
+            let next_hb = events
+                .next_map(TIMEOUT, |ev| match ev {
+                    Event::Sdam(hb @ SdamEvent::ServerHeartbeatStarted(_)) => Some(hb),
+                    Event::Sdam(hb @ SdamEvent::ServerHeartbeatFailed(_)) => Some(hb),
+                    Event::Sdam(hb @ SdamEvent::ServerHeartbeatSucceeded(_)) => Some(hb),
+                    _ => None,
+                })
+                .await
+                .unwrap();
+            // ... and see if it was a succeeded event.
+            if matches!(next_hb, SdamEvent::ServerHeartbeatSucceeded(_)) {
+                break;
+            }
+        }
+    })
+    .await
+    .unwrap();
+
+    // Send another ping
+    let mut events = c1.events.stream();
+    c1.database("admin")
+        .run_command(doc! { "ping": 1 })
+        .await
+        .unwrap();
+    let (start, _succeded) = events
+        .next_successful_command_execution(TIMEOUT, "ping")
+        .await
+        .unwrap();
+
+    // Assert that the cluster time hasn't changed
+    assert_eq!(cluster_time.as_ref(), start.command.get("$clusterTime"));
+}