diff --git a/CMakeLists.txt b/CMakeLists.txt
index 7ce9c42..e5876db 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -710,6 +710,7 @@
             src/armnnTfParser/test/TestMultiInputsOutputs.cpp
             src/armnnTfParser/test/Split.cpp
             src/armnnTfParser/test/Squeeze.cpp
+            src/armnnTfParser/test/Stack.cpp
             src/armnnTfParser/test/Sub.cpp
             src/armnnTfParser/test/StridedSlice.cpp
             )
diff --git a/src/armnnTfParser/TensorFlowSupport.md b/src/armnnTfParser/TensorFlowSupport.md
index 886dc7d..343d7a5 100644
--- a/src/armnnTfParser/TensorFlowSupport.md
+++ b/src/armnnTfParser/TensorFlowSupport.md
@@ -163,6 +163,10 @@
 
 The parser does not support all forms of broadcasting [broadcast composition](https://www.tensorflow.org/performance/xla/broadcasting), only broadcasting of scalars and 1D tensors. See the TensorFlow [subtract documentation](https://www.tensorflow.org/api_docs/python/tf/math/subtract) for more information.
 
+**stack**
+
+See the TensorFlow [stack documentation](https://www.tensorflow.org/api_docs/python/tf/stack) for more information.
+
 ## Tested networks
 
 Arm tests these operators with the following TensorFlow fp32 neural networks:
diff --git a/src/armnnTfParser/TfParser.cpp b/src/armnnTfParser/TfParser.cpp
index d65af23..b5a4211 100755
--- a/src/armnnTfParser/TfParser.cpp
+++ b/src/armnnTfParser/TfParser.cpp
@@ -376,7 +376,9 @@
     { "Minimum",               &TfParser::ParseMinimum },
     { "Equal",                 &TfParser::ParseEqual },
     { "Pad",                   &TfParser::ParsePad },
-    { "Sub",                   &TfParser::ParseSub }
+    { "Sub",                   &TfParser::ParseSub },
+    { "Pack" ,                 &TfParser::ParseStack },
+    { "Stack",                 &TfParser::ParseStack }
 };
 
 const std::list<std::string> TfParser::m_ControlInputs = {
@@ -1961,6 +1963,97 @@
     return std::make_unique<SingleLayerParsedTfOperation>(this, nodeDef, layer);
 }
 
+ParsedTfOperationPtr TfParser::ParseStack(const tensorflow::NodeDef& nodeDef, const tensorflow::GraphDef& graphDef)
+{
+    boost::ignore_unused(graphDef);
+    std::vector<OutputOfConstNodeDef> nodes = GetTfInputNodes(nodeDef);
+
+    unsigned int numInputs = static_cast<unsigned int>(nodes.size());
+    if (numInputs < 1)
+    {
+        throw ParseException(
+            boost::str(
+                boost::format(
+                    "Pack/Stack expects at least one input. Got %1% for Node %2% %3%")
+                % numInputs
+                % nodeDef.name()
+                % CHECK_LOCATION().AsString()));
+    }
+
+    std::vector<OutputOfParsedTfOperation> inputs = GetInputParsedTfOperationsChecked(nodeDef, numInputs);
+    // Use the tensor shape of the first input as the "correct" input shape in the descriptor
+    IOutputSlot* input0Slot = &inputs[0].m_IndexedValue->ResolveArmnnOutputSlot(inputs[0].m_Index);
+    const TensorInfo& inputTensorInfo = input0Slot->GetTensorInfo();
+    auto numDimensions = inputTensorInfo.GetShape().GetNumDimensions();
+
+    // validate axis
+    int32_t axis = ReadMandatoryNodeInt32Attribute(nodeDef, "axis");
+    const int sNumDimensions = (static_cast<int>(numDimensions) + 1);
+    if (!(axis < sNumDimensions && axis >= -sNumDimensions))
+    {
+        throw ParseException(
+            boost::str(
+                boost::format(
+                    "Axis index is not in range. Got %1% for Node %2% %3%")
+                % axis
+                % nodeDef.name()
+                % CHECK_LOCATION().AsString()));
+    }
+
+    if (axis < 0)
+    {
+        axis = static_cast<int32_t>(numDimensions) + axis + 1;
+    }
+
+    StackDescriptor stackDescriptor;
+    stackDescriptor.m_Axis = static_cast<uint32_t>(axis);
+    stackDescriptor.m_NumInputs = static_cast<uint32_t>(numInputs);
+    stackDescriptor.m_InputShape = inputTensorInfo.GetShape();
+
+    const unsigned int supportedNumDims = 4;
+    for (unsigned int viewIndex = 0; viewIndex < numInputs; ++viewIndex)
+    {
+        IOutputSlot& inputSlot = inputs[viewIndex].m_IndexedValue->ResolveArmnnOutputSlot(inputs[viewIndex].m_Index);
+        TensorInfo inputTensorInfo = inputSlot.GetTensorInfo();
+
+        // Double check dimensions of the tensors
+        if (inputTensorInfo.GetNumDimensions() >= supportedNumDims)
+        {
+            throw armnn::ParseException(
+                    boost::str(
+                        boost::format(
+                        "The number of dimensions: %1% for input tensors of the "
+                        "Pack/Stack op. Number of dimensions should be less than %2% %3%")
+                        % inputTensorInfo.GetNumDimensions()
+                        % supportedNumDims
+                        % CHECK_LOCATION().AsString()));
+        }
+    }
+
+    std::vector<unsigned int> outputDimensions;
+    for (unsigned int i = 0; i < stackDescriptor.m_InputShape.GetNumDimensions(); ++i)
+    {
+        outputDimensions.push_back(stackDescriptor.m_InputShape[i]);
+    }
+    outputDimensions.insert(outputDimensions.begin() + axis, numInputs);
+
+    // add Stack Layer
+    IConnectableLayer* const layer = m_Network->AddStackLayer(stackDescriptor, nodeDef.name().c_str());
+
+    for (unsigned int viewIndex = 0; viewIndex < numInputs; ++viewIndex)
+    {
+        IOutputSlot& inputSlot = inputs[viewIndex].m_IndexedValue->ResolveArmnnOutputSlot(inputs[viewIndex].m_Index);
+        inputSlot.Connect(layer->GetInputSlot(viewIndex));
+    }
+
+    layer->GetOutputSlot(0).SetTensorInfo(
+        armnn::TensorInfo(static_cast<uint32_t>(outputDimensions.size()),
+                          outputDimensions.data(),
+                          inputTensorInfo.GetDataType()));
+
+    return std::make_unique<SingleLayerParsedTfOperation>(this, nodeDef, layer);
+}
+
 unsigned int CheckPaddingTensor(const ConstTensor& paddingTensor,
                                 const TensorInfo& inputTensorInfo,
                                 const std::string& nodeName)
diff --git a/src/armnnTfParser/TfParser.hpp b/src/armnnTfParser/TfParser.hpp
index a7d02be..9277d44 100644
--- a/src/armnnTfParser/TfParser.hpp
+++ b/src/armnnTfParser/TfParser.hpp
@@ -170,6 +170,7 @@
     ParsedTfOperationPtr ParseGreater(const tensorflow::NodeDef& nodeDef, const tensorflow::GraphDef& graphDef);
     ParsedTfOperationPtr ParsePad(const tensorflow::NodeDef& nodeDef, const tensorflow::GraphDef& graphDef);
     ParsedTfOperationPtr ParseSub(const tensorflow::NodeDef& nodeDef, const tensorflow::GraphDef& graphDef);
+    ParsedTfOperationPtr ParseStack(const tensorflow::NodeDef& nodeDef, const tensorflow::GraphDef& graphDef);
     ParsedTfOperationPtr AddActivationLayer(const tensorflow::NodeDef& nodeDef, armnn::ActivationDescriptor& desc);
     ParsedTfOperationPtr AddAdditionLayer(const tensorflow::NodeDef& nodeDef, bool isBiasAdd = false);
     ParsedTfOperationPtr AddRealDivLayer(const tensorflow::NodeDef& nodeDef);
diff --git a/src/armnnTfParser/test/Stack.cpp b/src/armnnTfParser/test/Stack.cpp
new file mode 100644
index 0000000..b289917
--- /dev/null
+++ b/src/armnnTfParser/test/Stack.cpp
@@ -0,0 +1,174 @@
+//
+// Copyright © 2020 Arm Ltd. All rights reserved.
+// SPDX-License-Identifier: MIT
+//
+
+#include "armnnTfParser/ITfParser.hpp"
+#include "ParserPrototxtFixture.hpp"
+
+#include <PrototxtConversions.hpp>
+
+#include <boost/test/unit_test.hpp>
+
+BOOST_AUTO_TEST_SUITE(TensorflowParser)
+
+struct StackFixture : public armnnUtils::ParserPrototxtFixture<armnnTfParser::ITfParser>
+{
+    explicit StackFixture(const armnn::TensorShape& inputShape0,
+                          const armnn::TensorShape& inputShape1,
+                          int axis = 0)
+    {
+        m_Prototext = R"(
+        node {
+          name: "input0"
+          op: "Placeholder"
+          attr {
+            key: "dtype"
+            value {
+              type: DT_FLOAT
+            }
+          }
+          attr {
+            key: "shape"
+            value {
+              shape {
+              }
+            }
+          }
+        }
+        node {
+          name: "input1"
+          op: "Placeholder"
+          attr {
+            key: "dtype"
+            value {
+              type: DT_FLOAT
+            }
+          }
+          attr {
+            key: "shape"
+            value {
+              shape {
+              }
+            }
+          }
+        }
+        node {
+          name: "output"
+          op: "Stack"
+          input: "input0"
+          input: "input1"
+          attr {
+            key: "axis"
+            value {
+              i: )";
+                m_Prototext += std::to_string(axis);
+                m_Prototext += R"(
+            }
+          }
+        })";
+
+        Setup({{"input0", inputShape0 },
+               {"input1", inputShape1 }}, {"output"});
+    }
+};
+
+struct Stack3DFixture : StackFixture
+{
+    Stack3DFixture() : StackFixture({ 3, 2, 3 }, { 3, 2, 3 }, 3 ) {}
+};
+
+BOOST_FIXTURE_TEST_CASE(Stack3D, Stack3DFixture)
+{
+
+    RunTest<4>({ { "input0", {  1, 2, 3,
+                                4, 5, 6,
+
+                                7, 8, 9,
+                                10, 11, 12,
+
+                                13, 14, 15,
+                                16, 17, 18 } },
+                 { "input1", {  19, 20, 21,
+                                22, 23, 24,
+
+                                25, 26, 27,
+                                28, 29, 30,
+
+                                31, 32, 33,
+                                34, 35, 36 } } },
+               { { "output", {  1, 19,
+                                2, 20,
+                                3, 21,
+
+                                4, 22,
+                                5, 23,
+                                6, 24,
+
+                                7, 25,
+                                8, 26,
+                                9, 27,
+
+                                10, 28,
+                                11, 29,
+                                12, 30,
+
+                                13, 31,
+                                14, 32,
+                                15, 33,
+
+                                16, 34,
+                                17, 35,
+                                18, 36 } } });
+}
+
+struct Stack3DNegativeAxisFixture : StackFixture
+{
+    Stack3DNegativeAxisFixture() : StackFixture({ 3, 2, 3 }, { 3, 2, 3 }, -1 ) {}
+};
+
+BOOST_FIXTURE_TEST_CASE(Stack3DNegativeAxis, Stack3DNegativeAxisFixture)
+{
+
+    RunTest<4>({ { "input0", {  1, 2, 3,
+                                4, 5, 6,
+
+                                7, 8, 9,
+                                10, 11, 12,
+
+                                13, 14, 15,
+                                16, 17, 18 } },
+                 { "input1", {  19, 20, 21,
+                                22, 23, 24,
+
+                                25, 26, 27,
+                                28, 29, 30,
+
+                                31, 32, 33,
+                                34, 35, 36 } } },
+               { { "output", {  1, 19,
+                                2, 20,
+                                3, 21,
+
+                                4, 22,
+                                5, 23,
+                                6, 24,
+
+                                7, 25,
+                                8, 26,
+                                9, 27,
+
+                                10, 28,
+                                11, 29,
+                                12, 30,
+
+                                13, 31,
+                                14, 32,
+                                15, 33,
+
+                                16, 34,
+                                17, 35,
+                                18, 36 } } });
+}
+
+BOOST_AUTO_TEST_SUITE_END()