diff --git a/docs/OperatorKernels.md b/docs/OperatorKernels.md
index a9f57bf15e..b376ae03e9 100644
--- a/docs/OperatorKernels.md
+++ b/docs/OperatorKernels.md
@@ -1086,6 +1086,7 @@ Do not modify directly.*
 |GreaterOrEqual|*in* A:**T**<br> *in* B:**T**<br> *out* C:**T1**|16+|**T** = tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)<br/> **T1** = tensor(bool)|
 |||12+|**T** = tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)<br/> **T1** = tensor(bool)|
 |GridSample|*in* X:**T1**<br> *in* grid:**T2**<br> *out* Y:**T1**|16+|**T1** = tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8)<br/> **T2** = tensor(float), tensor(float16)|
+|GroupNorm||21+|**M** = tensor(float), tensor(float16)<br/> **T** = tensor(float), tensor(float16)|
 |HardSigmoid|*in* X:**T**<br> *out* Y:**T**|6+|**T** = tensor(float), tensor(float16)|
 |Hardmax|*in* input:**T**<br> *out* output:**T**|13+|**T** = tensor(float), tensor(float16)|
 |||11+|**T** = tensor(float), tensor(float16)|
diff --git a/onnxruntime/core/providers/dml/DmlExecutionProvider/src/Operators/OperatorRegistration.cpp b/onnxruntime/core/providers/dml/DmlExecutionProvider/src/Operators/OperatorRegistration.cpp
index fa5cbce2a9..b0b37d0137 100644
--- a/onnxruntime/core/providers/dml/DmlExecutionProvider/src/Operators/OperatorRegistration.cpp
+++ b/onnxruntime/core/providers/dml/DmlExecutionProvider/src/Operators/OperatorRegistration.cpp
@@ -1172,6 +1172,7 @@ constexpr static OperatorRegistrationInformation operatorRegistrationInformation
     {REG_INFO_MS(   1,  BiasAdd,                            typeNameListDefault,            supportedTypeListFloat16to32,           DmlGraphSupport::Supported)},
     {REG_INFO_MS(   1,  QuickGelu,                          typeNameListDefault,            supportedTypeListFloat16to32,           DmlGraphSupport::Supported)},
     {REG_INFO_MS(   1,  GroupNorm,                          typeNameListGroupNorm,          supportedTypeListGroupNorm,             DmlGraphSupport::Supported)},
+    {REG_INFO(      21, GroupNorm,                          typeNameListGroupNorm,          supportedTypeListGroupNorm,             DmlGraphSupport::Supported)},
     {REG_INFO_MS(   1,  MatMulNBits,                        typeNameListTwo,                supportedTypeListMatMulNBits,           DmlGraphSupport::Supported, requiredConstantCpuInputs(), std::nullopt, QueryMatMulNBits)},
 
     // Operators that need to alias an input with an output
diff --git a/onnxruntime/core/providers/dml/OperatorAuthorHelper/OperatorHelper.h b/onnxruntime/core/providers/dml/OperatorAuthorHelper/OperatorHelper.h
index 69b4225dee..c52e26dd32 100644
--- a/onnxruntime/core/providers/dml/OperatorAuthorHelper/OperatorHelper.h
+++ b/onnxruntime/core/providers/dml/OperatorAuthorHelper/OperatorHelper.h
@@ -1648,6 +1648,7 @@ using ShapeInferenceHelper_BatchNormalization15 = BatchNormalizationHelper;
 using ShapeInferenceHelper_LRN = GetOutputShapeAsInputShapeHelper;
 using ShapeInferenceHelper_MeanVarianceNormalization = GetOutputShapeAsInputShapeHelper;
 using ShapeInferenceHelper_GroupNorm = GetOutputShapeAsInputShapeHelper;
+using ShapeInferenceHelper_GroupNorm21 = GetOutputShapeAsInputShapeHelper;
 using ShapeInferenceHelper_LayerNormalization = GetOutputShapeAsInputShapeHelper;
 using ShapeInferenceHelper_LayerNormalization17 = GetOutputShapeAsInputShapeHelper;
 using ShapeInferenceHelper_SkipLayerNormalization = SkipLayerNormHelper;
diff --git a/onnxruntime/core/providers/dml/OperatorAuthorHelper/OperatorVersions.h b/onnxruntime/core/providers/dml/OperatorAuthorHelper/OperatorVersions.h
index c4f19271a1..b4d402a1d9 100644
--- a/onnxruntime/core/providers/dml/OperatorAuthorHelper/OperatorVersions.h
+++ b/onnxruntime/core/providers/dml/OperatorAuthorHelper/OperatorVersions.h
@@ -454,6 +454,7 @@ namespace OperatorHelper
         static const int sc_sinceVer_Transpose = 21;
         static const int sc_sinceVer_Identity = 21;
         static const int sc_sinceVer_QLinearMatMul = 21;
+        static const int sc_sinceVer_GroupNorm = 21;
     }
 
     namespace MsftOperatorSet1