diff --git a/CHANGELOG.md b/CHANGELOG.md
index b9b9675bf..22f5f773c 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -256,6 +256,11 @@
     * `WaitForSpineAnimationComplete` now proves an additional `bool includeEndEvent` parameter, defaults to `false` (previous behaviour).
     * Added a new `WaitForSpineAnimationEnd` yield instruction.
     * Added a new generic `WaitForSpineAnimation` yield instruction which can be configured to wait for any combination of animation track events. It is now used as base class for `WaitForSpineAnimationComplete` and `WaitForSpineAnimationEnd`.
+  * Additional **Fix Draw Order** parameter at SkeletonRenderer, defaults to `disabled` (previous behaviour).
+    Applies only when 3+ submeshes are used (2+ materials with alternating order, e.g. "A B A").
+		If true, MaterialPropertyBlocks are assigned at each material to prevent aggressive batching of submeshes
+		by e.g. the LWRP renderer, leading to incorrect draw order (e.g. "A1 B A2" changed to "A1A2 B").
+		You can leave this parameter disabled when everything is drawn correctly to save the additional performance cost.
 
 * **Changes of default values**
   * `SkeletonMecanim`'s `Layer Mix Mode` now defaults to `MixMode.SpineStyle` instead of `MixMode.MixAlways`.
diff --git a/spine-unity/Assets/Spine/Runtime/spine-unity/Components/SkeletonRenderer.cs b/spine-unity/Assets/Spine/Runtime/spine-unity/Components/SkeletonRenderer.cs
index e61906580..876709338 100644
--- a/spine-unity/Assets/Spine/Runtime/spine-unity/Components/SkeletonRenderer.cs
+++ b/spine-unity/Assets/Spine/Runtime/spine-unity/Components/SkeletonRenderer.cs
@@ -100,7 +100,7 @@ namespace Spine.Unity {
 		/// by e.g. the LWRP renderer, leading to incorrect draw order (e.g. "A1 B A2" changed to "A1A2 B").
 		/// You can disable this parameter when everything is drawn correctly to save the additional performance cost.
 		/// </summary>
-		public bool fixDrawOrder = true;
+		public bool fixDrawOrder = false;
 		#endif
 
 		/// <summary>If true, the mesh generator adds normals to the output mesh. For better performance and reduced memory requirements, use a shader that assumes the desired normal.</summary>