diff --git a/plugins/modules/ec2_instance.py b/plugins/modules/ec2_instance.py
index 22db3c88f79..e44bbfd3f2c 100644
--- a/plugins/modules/ec2_instance.py
+++ b/plugins/modules/ec2_instance.py
@@ -1443,16 +1443,19 @@ def get_default_subnet(ec2, vpc, availability_zone=None):
 
 
 def ensure_instance_state(state, ec2):
+    """
+    Sets return keys depending on the desired instance state
+    """
+    results = dict()
     if state in ('running', 'started'):
         changed, failed, instances, failure_reason = change_instance_state(filters=module.params.get('filters'), desired_state='RUNNING', ec2=ec2)
-
         if failed:
             module.fail_json(
                 msg="Unable to start instances: {0}".format(failure_reason),
                 reboot_success=list(changed),
                 reboot_failed=failed)
 
-        module.exit_json(
+        results = dict(
             msg='Instances started',
             reboot_success=list(changed),
             changed=bool(len(changed)),
@@ -1460,7 +1463,10 @@ def ensure_instance_state(state, ec2):
             instances=[pretty_instance(i) for i in instances],
         )
     elif state in ('restarted', 'rebooted'):
-        changed, failed, instances, failure_reason = change_instance_state(
+        # https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/ec2-instance-reboot.html
+        # The Ansible behaviour of issuing a stop/start has a minor impact on user billing
+        # This will need to be changelogged if we ever change to ec2.reboot_instance
+        _changed, _failed, _instances, _failure_reason = change_instance_state(
             filters=module.params.get('filters'),
             desired_state='STOPPED',
             ec2=ec2)
@@ -1475,7 +1481,7 @@ def ensure_instance_state(state, ec2):
                 reboot_success=list(changed),
                 reboot_failed=failed)
 
-        module.exit_json(
+        results = dict(
             msg='Instances restarted',
             reboot_success=list(changed),
             changed=bool(len(changed)),
@@ -1487,14 +1493,13 @@ def ensure_instance_state(state, ec2):
             filters=module.params.get('filters'),
             desired_state='STOPPED',
             ec2=ec2)
-
         if failed:
             module.fail_json(
                 msg="Unable to stop instances: {0}".format(failure_reason),
                 stop_success=list(changed),
                 stop_failed=failed)
 
-        module.exit_json(
+        results = dict(
             msg='Instances stopped',
             stop_success=list(changed),
             changed=bool(len(changed)),
@@ -1512,13 +1517,14 @@ def ensure_instance_state(state, ec2):
                 msg="Unable to terminate instances: {0}".format(failure_reason),
                 terminate_success=list(terminated),
                 terminate_failed=terminate_failed)
-        module.exit_json(
+        results = dict(
             msg='Instances terminated',
             terminate_success=list(terminated),
             changed=bool(len(terminated)),
             terminate_failed=[],
             instances=[pretty_instance(i) for i in instances],
         )
+    return results
 
 
 def change_instance_state(filters, desired_state, ec2):
@@ -1550,7 +1556,6 @@ def change_instance_state(filters, desired_state, ec2):
                 if module.check_mode:
                     changed.add(inst['InstanceId'])
                     continue
-
                 resp = ec2.stop_instances(aws_retry=True, InstanceIds=[inst['InstanceId']])
                 [changed.add(i['InstanceId']) for i in resp['StoppingInstances']]
             if desired_state == 'RUNNING':
@@ -1596,16 +1601,22 @@ def determine_iam_role(name_or_arn):
 
 
 def handle_existing(existing_matches, changed, ec2, state):
+    to_change_state = False
+    state_results = []
+    alter_config_result = []
     if state in ('running', 'started') and [i for i in existing_matches if i['State']['Name'] != 'running']:
-        ins_changed, failed, instances, failure_reason = change_instance_state(filters=module.params.get('filters'), desired_state='RUNNING', ec2=ec2)
-        if failed:
-            module.fail_json(msg="Couldn't start instances: {0}. Failure reason: {1}".format(instances, failure_reason))
-        module.exit_json(
-            changed=bool(len(ins_changed)) or changed,
-            instances=[pretty_instance(i) for i in instances],
-            instance_ids=[i['InstanceId'] for i in instances],
-        )
+        to_change_state = True
+    elif state == 'stopped' and [i for i in existing_matches if i['State']['Name'] != 'stopped']:
+        to_change_state = True
+    elif state in ('restarted', 'rebooted'):
+        to_change_state = True
+    if to_change_state:
+        # TODO: this breaks check_mode, because change_instance_state handles checkmode but not ensure
+        # because ensure passes off to change...
+        state_results = ensure_instance_state(state, ec2)
+
     changes = diff_instance_and_params(existing_matches[0], module.params, ec2)
+
     for c in changes:
         if not module.check_mode:
             try:
@@ -1616,24 +1627,21 @@ def handle_existing(existing_matches, changed, ec2, state):
     changed |= add_or_update_instance_profile(existing_matches[0], module.params.get('instance_role'), ec2)
     changed |= change_network_attachments(existing_matches[0], module.params, ec2)
     altered = find_instances(ec2, ids=[i['InstanceId'] for i in existing_matches])
-    module.exit_json(
+    alter_config_result = dict(
         changed=bool(len(changes)) or changed,
         instances=[pretty_instance(i) for i in altered],
         instance_ids=[i['InstanceId'] for i in altered],
         changes=changes,
     )
 
+    if len(state_results):
+        result = {**state_results, **alter_config_result}
+    else:
+        result = alter_config_result
+    return result
+
 
 def ensure_present(existing_matches, changed, ec2, state):
-    if len(existing_matches):
-        try:
-            handle_existing(existing_matches, changed, ec2, state)
-        except (botocore.exceptions.BotoCoreError, botocore.exceptions.ClientError) as e:
-            module.fail_json_aws(
-                e, msg="Failed to handle existing instances {0}".format(', '.join([i['InstanceId'] for i in existing_matches])),
-                # instances=[pretty_instance(i) for i in existing_matches],
-                # instance_ids=[i['InstanceId'] for i in existing_matches],
-            )
     try:
         instance_spec = build_run_instance_spec(module.params, ec2)
         # If check mode is enabled,suspend 'ensure function'.
@@ -1733,6 +1741,7 @@ def main():
         ],
         supports_check_mode=True
     )
+    result = dict()
 
     if module.params.get('network'):
         if module.params.get('network').get('interfaces'):
@@ -1748,10 +1757,6 @@ def main():
             # all states except shutting-down and terminated
             'instance-state-name': ['pending', 'running', 'stopping', 'stopped']
         }
-        if state == 'stopped':
-            # only need to change instances that aren't already stopped
-            filters['instance-state-name'] = ['stopping', 'pending', 'running']
-
         if isinstance(module.params.get('instance_ids'), string_types):
             filters['instance-id'] = [module.params.get('instance_ids')]
         elif isinstance(module.params.get('instance_ids'), list) and len(module.params.get('instance_ids')):
@@ -1799,11 +1804,15 @@ def main():
                 tags['Name'] = name
             changed |= manage_tags(match, tags, module.params.get('purge_tags', False), ec2)
 
-    if state in ('present', 'running', 'started'):
-        ensure_present(existing_matches=existing_matches, changed=changed, ec2=ec2, state=state)
-    elif state in ('restarted', 'rebooted', 'stopped', 'absent', 'terminated'):
+    if state in ('present', 'running', 'started', 'restarted', 'rebooted', 'stopped'):
+        # If we're not removing the instance, handle any config changes and update state
+        if len(existing_matches):
+            result = handle_existing(existing_matches, changed, ec2, state)
+        else:
+            ensure_present(existing_matches=existing_matches, changed=changed, ec2=ec2, state=state)
+    elif state in ('absent', 'terminated'):
         if existing_matches:
-            ensure_instance_state(state, ec2)
+            result = ensure_instance_state(state, ec2)
         else:
             module.exit_json(
                 msg='No matching instances found',
@@ -1813,6 +1822,8 @@ def main():
     else:
         module.fail_json(msg="We don't handle the state {0}".format(state))
 
+    module.exit_json(**result)
+
 
 if __name__ == '__main__':
     main()
diff --git a/tests/integration/targets/ec2_instance/inventory b/tests/integration/targets/ec2_instance/inventory
index a49c076d2f2..1513f3f40e4 100644
--- a/tests/integration/targets/ec2_instance/inventory
+++ b/tests/integration/targets/ec2_instance/inventory
@@ -12,6 +12,7 @@ termination_protection_wrapper
 tags_and_vpc_settings
 checkmode_tests
 security_group
+state_config_updates
 
 [all:vars]
 ansible_connection=local
diff --git a/tests/integration/targets/ec2_instance/roles/ec2_instance/tasks/state_config_updates.yml b/tests/integration/targets/ec2_instance/roles/ec2_instance/tasks/state_config_updates.yml
new file mode 100644
index 00000000000..6b3f524b870
--- /dev/null
+++ b/tests/integration/targets/ec2_instance/roles/ec2_instance/tasks/state_config_updates.yml
@@ -0,0 +1,145 @@
+# Test that configuration changes, like security groups and instance attributes,
+# are updated correctly when the instance has different states, and also when
+# changing the state of an instance.
+# https://github.com/ansible-collections/community.aws/issues/16
+- block:
+  - name: "Make instance with sg and termination protection enabled"
+    ec2_instance:
+      state: present
+      name: "{{ resource_prefix }}-test-state-param-changes"
+      image_id: "{{ ec2_ami_image }}"
+      tags:
+        TestId: "{{ ec2_instance_tag_TestId }}"
+      security_groups: "{{ sg.group_id }}"
+      vpc_subnet_id: "{{ testing_subnet_b.subnet.id }}"
+      termination_protection: False
+      instance_type: "{{ ec2_instance_type }}"
+      wait: True
+    register: create_result
+
+  - assert:
+      that:
+        - create_result is not failed
+        - create_result.changed
+        - '"instances" in create_result'
+        - '"instance_ids" in create_result'
+        - '"spec" in create_result'
+        - create_result.instances[0].security_groups[0].group_id == "{{ sg.group_id }}"
+        - create_result.spec.DisableApiTermination == False
+
+  - name: "Change sg and termination protection while instance is in state running"
+    ec2_instance:
+      state: running
+      name: "{{ resource_prefix }}-test-state-param-changes"
+      image_id: "{{ ec2_ami_image }}"
+      tags:
+        TestId: "{{ ec2_instance_tag_TestId }}"
+      security_groups: "{{ sg2.group_id }}"
+      vpc_subnet_id: "{{ testing_subnet_b.subnet.id }}"
+      termination_protection: True
+      instance_type: "{{ ec2_instance_type }}"
+    register: change_params_result
+
+  - assert:
+      that:
+        - change_params_result is not failed
+        - change_params_result.changed
+        - '"instances" in change_params_result'
+        - '"instance_ids" in change_params_result'
+        - '"changes" in change_params_result'
+        - change_params_result.instances[0].security_groups[0].group_id == "{{ sg2.group_id }}"
+        - change_params_result.changes[0].DisableApiTermination.Value == True
+        - change_params_result.changes[1].Groups[0] == "{{ sg2.group_id }}"  # TODO fix this to be less fragile
+
+
+  - name: "Change instance state from running to stopped, and change sg and termination protection"
+    ec2_instance:
+      state: stopped
+      name: "{{ resource_prefix }}-test-state-param-changes"
+      image_id: "{{ ec2_ami_image }}"
+      tags:
+        TestId: "{{ ec2_instance_tag_TestId }}"
+      security_groups: "{{ sg.group_id }}"
+      vpc_subnet_id: "{{ testing_subnet_b.subnet.id }}"
+      termination_protection: False
+      instance_type: "{{ ec2_instance_type }}"
+    register: change_state_params_result
+
+  - assert:
+      that:
+        - change_state_params_result is not failed
+        - change_state_params_result.changed
+        - '"instances" in change_state_params_result'
+        - '"instance_ids" in change_state_params_result'
+        - '"changes" in change_state_params_result'
+        - '"stop_success" in change_state_params_result'
+        - '"stop_failed" in change_state_params_result'
+        - change_state_params_result.instances[0].security_groups[0].group_id == "{{ sg.group_id }}"
+        - change_state_params_result.changes[0].DisableApiTermination.Value == False
+
+  - name: "Change sg and termination protection while instance is in state stopped"
+    ec2_instance:
+      state: stopped
+      name: "{{ resource_prefix }}-test-state-param-changes"
+      image_id: "{{ ec2_ami_image }}"
+      tags:
+        TestId: "{{ ec2_instance_tag_TestId }}"
+      security_groups: "{{ sg2.group_id }}"
+      vpc_subnet_id: "{{ testing_subnet_b.subnet.id }}"
+      termination_protection: True
+      instance_type: "{{ ec2_instance_type }}"
+    register: change_params_stopped_result
+
+  - assert:
+      that:
+        - change_params_stopped_result is not failed
+        - change_params_stopped_result.changed
+        - '"instances" in change_params_stopped_result'
+        - '"instance_ids" in change_params_stopped_result'
+        - '"changes" in change_params_stopped_result'
+        - change_params_stopped_result.instances[0].security_groups[0].group_id == "{{ sg2.group_id }}"
+        - change_params_stopped_result.changes[0].DisableApiTermination.Value == True
+
+  - name: "Change instance state from stopped to running, and change sg and termination protection"
+    ec2_instance:
+      state: running
+      name: "{{ resource_prefix }}-test-state-param-changes"
+      image_id: "{{ ec2_ami_image }}"
+      tags:
+        TestId: "{{ ec2_instance_tag_TestId }}"
+      security_groups: "{{ sg.group_id }}"
+      vpc_subnet_id: "{{ testing_subnet_b.subnet.id }}"
+      termination_protection: False
+      instance_type: "{{ ec2_instance_type }}"
+    register: change_params_start_result
+
+  - assert:
+      that:
+        - change_params_start_result is not failed
+        - change_params_start_result.changed
+        - '"instances" in change_params_start_result'
+        - '"instance_ids" in change_params_start_result'
+        - '"changes" in change_params_start_result'
+        - '"reboot_success" in change_params_start_result'
+        - '"reboot_failed" in change_params_start_result'
+        - change_params_start_result.instances[0].security_groups[0].group_id == "{{ sg.group_id }}"
+        - change_params_start_result.changes[0].DisableApiTermination.Value == False
+
+  always:
+
+    - name: Set termination protection to false (so we can terminate instance) (cleanup)
+      ec2_instance:
+        filters:
+          tag:TestId: "{{ ec2_instance_tag_TestId }}"
+        termination_protection: False
+      ignore_errors: yes
+
+    - name: Terminate instance
+      ec2_instance:
+        filters:
+          tag:TestId: "{{ ec2_instance_tag_TestId }}"
+        state: absent
+        wait: False
+      ignore_errors: yes
+
+    - include_tasks: env_cleanup.yml
\ No newline at end of file