diff --git a/clients/src/main/java/org/apache/kafka/clients/consumer/internals/AsyncKafkaConsumer.java b/clients/src/main/java/org/apache/kafka/clients/consumer/internals/AsyncKafkaConsumer.java
index ec27794e6d938..38074ebb7aecb 100644
--- a/clients/src/main/java/org/apache/kafka/clients/consumer/internals/AsyncKafkaConsumer.java
+++ b/clients/src/main/java/org/apache/kafka/clients/consumer/internals/AsyncKafkaConsumer.java
@@ -1990,6 +1990,13 @@ private Fetch<K, V> pollForFetches(Timer timer) {
             }
         }
 
+        // If the background thread has already discovered a metadata error (e.g. TopicAuthorizationException)
+        // and completed the inflight poll event, skip blocking on the fetch buffer entirely. The error will
+        // be surfaced by checkInflightPoll on the next iteration of the poll loop. See KAFKA-20397.
+        if (inflightPoll != null && inflightPoll.error().isPresent()) {
+            return collectFetch();
+        }
+
         log.trace("Polling for fetches with timeout {}", pollTimeout);
 
         Timer pollTimer = time.timer(pollTimeout);
diff --git a/clients/src/test/java/org/apache/kafka/clients/consumer/internals/AsyncKafkaConsumerTest.java b/clients/src/test/java/org/apache/kafka/clients/consumer/internals/AsyncKafkaConsumerTest.java
index 9743da654b258..0623550d37972 100644
--- a/clients/src/test/java/org/apache/kafka/clients/consumer/internals/AsyncKafkaConsumerTest.java
+++ b/clients/src/test/java/org/apache/kafka/clients/consumer/internals/AsyncKafkaConsumerTest.java
@@ -71,6 +71,7 @@
 import org.apache.kafka.common.errors.GroupAuthorizationException;
 import org.apache.kafka.common.errors.InterruptException;
 import org.apache.kafka.common.errors.InvalidGroupIdException;
+import org.apache.kafka.common.errors.TopicAuthorizationException;
 import org.apache.kafka.common.errors.TimeoutException;
 import org.apache.kafka.common.errors.UnsupportedVersionException;
 import org.apache.kafka.common.errors.WakeupException;
@@ -1736,6 +1737,73 @@ public void testPollDoesNotAddNewAsyncPollEventWhenOneIsAlreadyInFlight() {
         verify(applicationEventHandler, times(1)).add(isA(AsyncPollEvent.class));
     }
 
+    /**
+     * KAFKA-20397: verifies that a metadata error discovered by the background thread between
+     * {@code checkInflightPoll()} and the blocking wait in {@code pollForFetches()} is surfaced
+     * promptly, without wasting the full fetch-wait interval.
+     *
+     * <p>Scenario: {@code maximumTimeToWait()} returns 100 ms (shorter than the 500 ms poll
+     * timeout).  The background thread completes the inflight {@link AsyncPollEvent} with a
+     * {@link TopicAuthorizationException} before {@code pollForFetches()} reaches the blocking
+     * {@code fetchBuffer.awaitWakeup()} call.
+     */
+    @Test
+    public void testPollSurfacesMetadataErrorWithoutWastingFetchWaitInterval() {
+        FetchBuffer fetchBuffer = mock(FetchBuffer.class);
+        ConsumerInterceptors<String, String> interceptors = mock(ConsumerInterceptors.class);
+        ConsumerRebalanceListenerInvoker rebalanceListenerInvoker = mock(ConsumerRebalanceListenerInvoker.class);
+        SubscriptionState subscriptions = new SubscriptionState(new LogContext(), AutoOffsetResetStrategy.NONE);
+        consumer = newConsumer(fetchBuffer, interceptors, rebalanceListenerInvoker, subscriptions);
+
+        final String topicName = "topic1";
+        final TopicPartition tp = new TopicPartition(topicName, 0);
+
+        subscriptions.assignFromUser(singleton(tp));
+        subscriptions.seek(tp, 0);
+
+        final long fetchWaitMs = 100L;
+        doReturn(Fetch.empty()).when(fetchCollector).collectFetch(any(FetchBuffer.class));
+        doReturn(LeaderAndEpoch.noLeaderOrEpoch()).when(metadata).currentLeader(any());
+
+        AtomicReference<AsyncPollEvent> capturedEvent = new AtomicReference<>();
+        doAnswer(invocation -> {
+            capturedEvent.set(invocation.getArgument(0));
+            return null;
+        }).when(applicationEventHandler).add(ArgumentMatchers.isA(AsyncPollEvent.class));
+
+        AtomicBoolean errorInjected = new AtomicBoolean(false);
+
+        // Inject the error inside maximumTimeToWait(), which is the first call in
+        // pollForFetches — after checkInflightPoll has already submitted the event.
+        // This models the background thread completing the event with a metadata error
+        // between checkInflightPoll and the blocking wait.
+        doAnswer(invocation -> {
+            if (!errorInjected.get() && capturedEvent.get() != null) {
+                capturedEvent.get().onMetadataError(
+                    new TopicAuthorizationException(singleton(topicName)));
+                errorInjected.set(true);
+            }
+            return fetchWaitMs;
+        }).when(applicationEventHandler).maximumTimeToWait();
+
+        doAnswer(invocation -> {
+            Timer pollTimer = invocation.getArgument(0, Timer.class);
+            ((MockTime) time).sleep(pollTimer.remainingMs());
+            return null;
+        }).when(fetchBuffer).awaitWakeup(any(Timer.class));
+
+        final long pollTimeoutMs = 500;
+        long startMs = time.milliseconds();
+
+        assertThrows(TopicAuthorizationException.class,
+            () -> consumer.poll(Duration.ofMillis(pollTimeoutMs)));
+        long elapsedMs = time.milliseconds() - startMs;
+
+        assertTrue(elapsedMs < fetchWaitMs,
+            "Expected error to be surfaced promptly (elapsed " + elapsedMs +
+            " ms), but consumer wasted a full fetch-wait interval (" + fetchWaitMs + " ms)");
+    }
+
     /**
      * Tests {@link AsyncKafkaConsumer#processBackgroundEvents(Future, Timer, Predicate) processBackgroundEvents}
      * handles the case where the {@link Future} takes a bit of time to complete, but does within the timeout.