diff --git a/src/armnnTfLiteParser/TfLiteParser.cpp b/src/armnnTfLiteParser/TfLiteParser.cpp
index d1d45f5..c3d56b1 100644
--- a/src/armnnTfLiteParser/TfLiteParser.cpp
+++ b/src/armnnTfLiteParser/TfLiteParser.cpp
@@ -1845,6 +1845,8 @@
     TfLiteParser::TensorRawPtrVector outputs = GetOutputs(m_Model, subgraphIndex, operatorIndex);
     CHECK_VALID_SIZE(outputs.size(), 1);
 
+    armnn::TensorInfo inputTensorInfo = ToTensorInfo(inputs[0]);
+
     armnn::TensorInfo padTensorInfo = ToTensorInfo(inputs[1]);
     BufferRawPtr bufferPtr = GetBuffer(m_Model, inputs[1]->buffer);
 
@@ -1853,6 +1855,10 @@
 
     size_t step = 2;
     armnn::PadDescriptor desc;
+    if (inputTensorInfo.IsQuantized())
+    {
+        desc.m_PadValue = static_cast<float>(inputTensorInfo.GetQuantizationOffset());
+    }
     for (unsigned int i = 0; i < padTensorInfo.GetNumElements() / step; ++i)
     {
         desc.m_PadList.emplace_back(padBuffer[i * step], padBuffer[i * step + 1]);
diff --git a/src/armnnTfLiteParser/test/Pad.cpp b/src/armnnTfLiteParser/test/Pad.cpp
index bdc8478..aab1536 100644
--- a/src/armnnTfLiteParser/test/Pad.cpp
+++ b/src/armnnTfLiteParser/test/Pad.cpp
@@ -14,10 +14,13 @@
 
 struct PadFixture : public ParserFlatbuffersFixture
 {
-    explicit PadFixture(const std::string & inputShape,
-                        const std::string & outputShape,
-                        const std::string & padListShape,
-                        const std::string & padListData)
+    explicit PadFixture(const std::string& inputShape,
+                        const std::string& outputShape,
+                        const std::string& padListShape,
+                        const std::string& padListData,
+                        const std::string& dataType = "FLOAT32",
+                        const std::string& scale = "1.0",
+                        const std::string& offset = "0")
     {
         m_JsonString = R"(
             {
@@ -27,26 +30,26 @@
                     "tensors": [
                         {
                             "shape": )" + inputShape + R"(,
-                            "type": "FLOAT32",
+                            "type": )" + dataType + R"(,
                             "buffer": 0,
                             "name": "inputTensor",
                             "quantization": {
                                 "min": [ 0.0 ],
                                 "max": [ 255.0 ],
-                                "scale": [ 1.0 ],
-                                "zero_point": [ 0 ],
+                                "scale": [ )" + scale + R"( ],
+                                "zero_point": [ )" + offset + R"( ],
                             }
                         },
                         {
                              "shape": )" + outputShape + R"(,
-                             "type": "FLOAT32",
+                             "type": )" + dataType + R"(,
                              "buffer": 1,
                              "name": "outputTensor",
                              "quantization": {
                                 "min": [ 0.0 ],
                                 "max": [ 255.0 ],
-                                "scale": [ 1.0 ],
-                                "zero_point": [ 0 ],
+                                "scale": [ )" + scale + R"( ],
+                                "zero_point": [ )" + offset + R"( ],
                             }
                         },
                         {
@@ -101,4 +104,40 @@
                               0.0f, 0.0f, 0.0f, 0.0f, 0.0f, 0.0f, 0.0f }}});
 }
 
+struct Uint8PadFixture : public PadFixture
+{
+    Uint8PadFixture() : PadFixture("[ 2, 3 ]", "[ 4, 7 ]", "[ 2, 2 ]",
+                                  "[  1,0,0,0, 1,0,0,0, 2,0,0,0, 2,0,0,0 ]",
+                                  "UINT8", "-2.0", "3") {}
+};
+
+BOOST_FIXTURE_TEST_CASE(ParsePadUint8, Uint8PadFixture)
+{
+    RunTest<2, armnn::DataType::QAsymmU8>
+        (0,
+         {{ "inputTensor",  { 1, 2, 3, 4, 5, 6 }}},
+         {{ "outputTensor", { 3, 3, 3, 3, 3, 3, 3,
+                              3, 3, 1, 2, 3, 3, 3,
+                              3, 3, 4, 5, 6, 3, 3,
+                              3, 3, 3, 3, 3, 3, 3 }}});
+}
+
+struct Int8PadFixture : public PadFixture
+{
+    Int8PadFixture() : PadFixture("[ 2, 3 ]", "[ 4, 7 ]", "[ 2, 2 ]",
+                                    "[  1,0,0,0, 1,0,0,0, 2,0,0,0, 2,0,0,0 ]",
+                                    "INT8", "-2.0", "3") {}
+};
+
+BOOST_FIXTURE_TEST_CASE(ParsePadInt8, Int8PadFixture)
+{
+    RunTest<2, armnn::DataType::QAsymmS8>
+        (0,
+         {{ "inputTensor",  { 1, -2, 3, 4, 5, -6 }}},
+         {{ "outputTensor", { 3, 3, 3, 3, 3, 3, 3,
+                              3, 3, 1, -2, 3, 3, 3,
+                              3, 3, 4, 5, -6, 3, 3,
+                              3, 3, 3, 3, 3, 3, 3 }}});
+}
+
 BOOST_AUTO_TEST_SUITE_END()
diff --git a/src/backends/backendsCommon/test/layerTests/PadTestImpl.cpp b/src/backends/backendsCommon/test/layerTests/PadTestImpl.cpp
index 5511f86..086f875 100644
--- a/src/backends/backendsCommon/test/layerTests/PadTestImpl.cpp
+++ b/src/backends/backendsCommon/test/layerTests/PadTestImpl.cpp
@@ -417,6 +417,79 @@
     return result;
 }
 
+template<armnn::DataType ArmnnType, typename T>
+LayerTestResult<T, 2> PadQAsymmTestCommon(
+    armnn::IWorkloadFactory& workloadFactory,
+    const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager,
+    const armnn::ITensorHandleFactory& tensorHandleFactory,
+    float qScale,
+    int32_t qOffset,
+    const float customPaddingValue)
+{
+    IgnoreUnused(memoryManager);
+    const armnn::TensorShape inputShape{ 3, 3 };
+    const armnn::TensorShape outputShape{ 7, 7 };
+
+    const armnn::TensorInfo inputTensorInfo(inputShape, ArmnnType, qScale, qOffset);
+    const armnn::TensorInfo outputTensorInfo(outputShape, ArmnnType, qScale, qOffset);
+
+    std::vector<T> inputValues =
+        {
+            // Height (3) x Width (3)
+            4, 8, 6,
+            7, 4, 4,
+            3, 2, 4
+        };
+
+    T p = static_cast<T>(customPaddingValue);
+    std::vector<T> expectedOutputValues =
+        {
+            p, p, p, p, p, p, p,
+            p, p, p, p, p, p, p,
+            p, p, 4, 8, 6, p, p,
+            p, p, 7, 4, 4, p, p,
+            p, p, 3, 2, 4, p, p,
+            p, p, p, p, p, p, p,
+            p, p, p, p, p, p, p
+        };
+
+    auto inputTensor = MakeTensor<T, 2>(inputTensorInfo, std::vector<T>(inputValues));
+
+    LayerTestResult<T, 2> result(outputTensorInfo);
+    result.outputExpected = MakeTensor<T, 2>(outputTensorInfo, std::vector<T>(expectedOutputValues));
+
+    std::unique_ptr<armnn::ITensorHandle> inputHandle = tensorHandleFactory.CreateTensorHandle(inputTensorInfo);
+    std::unique_ptr<armnn::ITensorHandle> outputHandle = tensorHandleFactory.CreateTensorHandle(outputTensorInfo);
+
+
+    armnn::PadQueueDescriptor descriptor;
+
+    std::vector<std::pair<unsigned int, unsigned int>> padList;
+    padList.push_back(std::pair<unsigned int, unsigned int>(2,2));
+    padList.push_back(std::pair<unsigned int, unsigned int>(2,2));
+
+    descriptor.m_Parameters.m_PadList = padList;
+    descriptor.m_Parameters.m_PadValue = customPaddingValue;
+    armnn::WorkloadInfo info;
+
+    AddInputToWorkload(descriptor, info, inputTensorInfo, inputHandle.get());
+    AddOutputToWorkload(descriptor, info, outputTensorInfo, outputHandle.get());
+
+    std::unique_ptr<armnn::IWorkload> workload = workloadFactory.CreatePad(descriptor, info);
+
+    inputHandle->Allocate();
+    outputHandle->Allocate();
+
+    CopyDataToITensorHandle(inputHandle.get(), &inputTensor[0][0]);
+
+    workload->PostAllocationConfigure();
+    workload->Execute();
+
+    CopyDataFromITensorHandle(&result.output[0][0], outputHandle.get());
+
+    return result;
+}
+
 //
 // Explicit template specializations
 //
@@ -446,6 +519,24 @@
     float qScale,
     int32_t qOffset);
 
+template LayerTestResult<armnn::ResolveType<armnn::DataType::QAsymmS8>, 2>
+PadQAsymmTestCommon<armnn::DataType::QAsymmS8>(
+    armnn::IWorkloadFactory& workloadFactory,
+    const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager,
+    const armnn::ITensorHandleFactory& tensorHandleFactory,
+    float qScale,
+    int32_t qOffset,
+    const float customPaddingValue);
+
+template LayerTestResult<armnn::ResolveType<armnn::DataType::QAsymmU8>, 2>
+PadQAsymmTestCommon<armnn::DataType::QAsymmU8>(
+    armnn::IWorkloadFactory& workloadFactory,
+    const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager,
+    const armnn::ITensorHandleFactory& tensorHandleFactory,
+    float qScale,
+    int32_t qOffset,
+    const float customPaddingValue);
+
 //
 // Implementation functions
 //
@@ -582,3 +673,21 @@
 {
     return Pad4dTestCommon<armnn::DataType::QSymmS8>(workloadFactory, memoryManager, tensorHandleFactory, 1.0f, 0);
 }
+
+LayerTestResult<int8_t, 2> PadInt8AsymmTest(
+    armnn::IWorkloadFactory& workloadFactory,
+    const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager,
+    const armnn::ITensorHandleFactory& tensorHandleFactory)
+{
+    return PadQAsymmTestCommon<armnn::DataType::QAsymmS8>(
+        workloadFactory, memoryManager, tensorHandleFactory, 2.0f, 2);
+}
+
+LayerTestResult<int8_t, 2> PadInt8CustomPaddingAsymmTest(
+    armnn::IWorkloadFactory& workloadFactory,
+    const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager,
+    const armnn::ITensorHandleFactory& tensorHandleFactory)
+{
+    return PadQAsymmTestCommon<armnn::DataType::QAsymmS8>(
+        workloadFactory, memoryManager, tensorHandleFactory, 2.0f, 3, 1.0f);
+}
diff --git a/src/backends/backendsCommon/test/layerTests/PadTestImpl.hpp b/src/backends/backendsCommon/test/layerTests/PadTestImpl.hpp
index 8d60386..4c30c42 100644
--- a/src/backends/backendsCommon/test/layerTests/PadTestImpl.hpp
+++ b/src/backends/backendsCommon/test/layerTests/PadTestImpl.hpp
@@ -39,6 +39,15 @@
     float qScale,
     int32_t qOffset);
 
+template<armnn::DataType ArmnnType, typename T = armnn::ResolveType<ArmnnType>>
+LayerTestResult<T, 2> PadQAsymmTestCommon(
+    armnn::IWorkloadFactory& workloadFactory,
+    const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager,
+    const armnn::ITensorHandleFactory& tensorHandleFactory,
+    float qScale,
+    int32_t qOffset,
+    const float customPaddingValue = 0.0f);
+
 LayerTestResult<uint8_t, 2> PadUint82dTest(
     armnn::IWorkloadFactory& workloadFactory,
     const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager,
@@ -117,4 +126,14 @@
 LayerTestResult<int8_t, 4> PadInt84dTest(
     armnn::IWorkloadFactory& workloadFactory,
     const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager,
+    const armnn::ITensorHandleFactory& tensorHandleFactory);
+
+LayerTestResult<int8_t, 2> PadInt82dAsymmTest(
+    armnn::IWorkloadFactory& workloadFactory,
+    const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager,
+    const armnn::ITensorHandleFactory& tensorHandleFactory);
+
+LayerTestResult<int8_t, 2> PadInt82dCustomPaddingAsymmTest(
+    armnn::IWorkloadFactory& workloadFactory,
+    const armnn::IBackendInternal::IMemoryManagerSharedPtr& memoryManager,
     const armnn::ITensorHandleFactory& tensorHandleFactory);
\ No newline at end of file
diff --git a/src/backends/reference/test/RefLayerTests.cpp b/src/backends/reference/test/RefLayerTests.cpp
index be95ad7..502e0cb 100644
--- a/src/backends/reference/test/RefLayerTests.cpp
+++ b/src/backends/reference/test/RefLayerTests.cpp
@@ -1309,6 +1309,8 @@
 ARMNN_AUTO_TEST_CASE_WITH_THF(PadUint82dCustomPadding, PadUint82dCustomPaddingTest)
 ARMNN_AUTO_TEST_CASE_WITH_THF(PadUint83d, PadUint83dTest)
 ARMNN_AUTO_TEST_CASE_WITH_THF(PadUint84d, PadUint84dTest)
+ARMNN_AUTO_TEST_CASE_WITH_THF(PadUint8Asymm, PadQAsymmTestCommon<DataType::QAsymmU8>, -2.0f, 3, 0.0f)
+ARMNN_AUTO_TEST_CASE_WITH_THF(PadUint8CustomPaddingAsymm, PadQAsymmTestCommon<DataType::QAsymmU8>, -2.0f, 3, 2.0f)
 
 ARMNN_AUTO_TEST_CASE_WITH_THF(Pad2dQSymm16, Pad2dTestCommon<DataType::QSymmS16>, 2.0f, 0, 0.0f)
 ARMNN_AUTO_TEST_CASE_WITH_THF(Pad2dQSymm16CustomPadding, Pad2dTestCommon<DataType::QSymmS16>, 2.0f, 0, 1.0f)
@@ -1319,6 +1321,8 @@
 ARMNN_AUTO_TEST_CASE_WITH_THF(PadInt82dCustomPadding, PadInt82dCustomPaddingTest)
 ARMNN_AUTO_TEST_CASE_WITH_THF(PadInt83d, PadInt83dTest)
 ARMNN_AUTO_TEST_CASE_WITH_THF(PadInt84d, PadInt84dTest)
+ARMNN_AUTO_TEST_CASE_WITH_THF(PadQAsymmS8, PadQAsymmTestCommon<DataType::QAsymmS8>, -2.0f, 3, 0.0f)
+ARMNN_AUTO_TEST_CASE_WITH_THF(PadQAsymmS8CustomPadding, PadQAsymmTestCommon<DataType::QAsymmS8>, -2.0f, 3, 2.0f)
 
 // Constant
 ARMNN_AUTO_TEST_CASE_WITH_THF(Constant, ConstantTest)