diff --git a/moto/autoscaling/models.py b/moto/autoscaling/models.py
index 45ee7d192..88577433e 100644
--- a/moto/autoscaling/models.py
+++ b/moto/autoscaling/models.py
@@ -655,10 +655,16 @@ class AutoScalingBackend(BaseBackend):
         self.set_desired_capacity(group_name, 0)
         self.autoscaling_groups.pop(group_name, None)
 
-    def describe_auto_scaling_instances(self):
+    def describe_auto_scaling_instances(self, instance_ids):
         instance_states = []
         for group in self.autoscaling_groups.values():
-            instance_states.extend(group.instance_states)
+            instance_states.extend(
+                [
+                    x
+                    for x in group.instance_states
+                    if not instance_ids or x.instance.id in instance_ids
+                ]
+            )
         return instance_states
 
     def attach_instances(self, group_name, instance_ids):
diff --git a/moto/autoscaling/responses.py b/moto/autoscaling/responses.py
index 83e2f7d5a..41c79edb4 100644
--- a/moto/autoscaling/responses.py
+++ b/moto/autoscaling/responses.py
@@ -226,7 +226,9 @@ class AutoScalingResponse(BaseResponse):
         return template.render()
 
     def describe_auto_scaling_instances(self):
-        instance_states = self.autoscaling_backend.describe_auto_scaling_instances()
+        instance_states = self.autoscaling_backend.describe_auto_scaling_instances(
+            instance_ids=self._get_multi_param("InstanceIds.member")
+        )
         template = self.response_template(DESCRIBE_AUTOSCALING_INSTANCES_TEMPLATE)
         return template.render(instance_states=instance_states)
 
diff --git a/tests/test_autoscaling/test_autoscaling.py b/tests/test_autoscaling/test_autoscaling.py
index 2e7255381..5cf3dc6ff 100644
--- a/tests/test_autoscaling/test_autoscaling.py
+++ b/tests/test_autoscaling/test_autoscaling.py
@@ -843,13 +843,41 @@ def test_describe_autoscaling_instances_boto3():
         NewInstancesProtectedFromScaleIn=True,
     )
 
+    response = client.describe_auto_scaling_instances()
+    len(response["AutoScalingInstances"]).should.equal(5)
+    for instance in response["AutoScalingInstances"]:
+        instance["AutoScalingGroupName"].should.equal("test_asg")
+        instance["AvailabilityZone"].should.equal("us-east-1a")
+        instance["ProtectedFromScaleIn"].should.equal(True)
+
+
+@mock_autoscaling
+def test_describe_autoscaling_instances_instanceid_filter():
+    mocked_networking = setup_networking()
+    client = boto3.client("autoscaling", region_name="us-east-1")
+    _ = client.create_launch_configuration(
+        LaunchConfigurationName="test_launch_configuration"
+    )
+    _ = client.create_auto_scaling_group(
+        AutoScalingGroupName="test_asg",
+        LaunchConfigurationName="test_launch_configuration",
+        MinSize=0,
+        MaxSize=20,
+        DesiredCapacity=5,
+        VPCZoneIdentifier=mocked_networking["subnet1"],
+        NewInstancesProtectedFromScaleIn=True,
+    )
+
     response = client.describe_auto_scaling_groups(AutoScalingGroupNames=["test_asg"])
     instance_ids = [
         instance["InstanceId"]
         for instance in response["AutoScalingGroups"][0]["Instances"]
     ]
 
-    response = client.describe_auto_scaling_instances(InstanceIds=instance_ids)
+    response = client.describe_auto_scaling_instances(
+        InstanceIds=instance_ids[0:2]
+    )  # Filter by first 2 of 5
+    len(response["AutoScalingInstances"]).should.equal(2)
     for instance in response["AutoScalingInstances"]:
         instance["AutoScalingGroupName"].should.equal("test_asg")
         instance["AvailabilityZone"].should.equal("us-east-1a")