diff --git a/src/backends/gpuFsa/test/GpuFsaEndToEndTests.cpp b/src/backends/gpuFsa/test/GpuFsaEndToEndTests.cpp
index 79dd9d3..9972b24 100644
--- a/src/backends/gpuFsa/test/GpuFsaEndToEndTests.cpp
+++ b/src/backends/gpuFsa/test/GpuFsaEndToEndTests.cpp
@@ -6,7 +6,9 @@
 #include "backendsCommon/test/EndToEndTestImpl.hpp"
 
 #include "backendsCommon/test/Convolution2dEndToEndTestImpl.hpp"
+
 #include "backendsCommon/test/DepthwiseConvolution2dEndToEndTests.hpp"
+#include "backendsCommon/test/ElementwiseBinaryEndToEndTestImpl.hpp"
 
 #include <doctest/doctest.h>
 
@@ -32,4 +34,15 @@
                                                                                        armnn::DataLayout::NHWC);
 }
 
+// ElementwiseBinary Add
+TEST_CASE("GpuFsaElementwiseBinaryAddTestFloat32")
+{
+    ElementwiseBinarySimple3DEndToEnd<armnn::DataType::Float32>(gpuFsaDefaultBackends, BinaryOperation::Add);
+}
+
+TEST_CASE("GpuFsaElementwiseBinaryAddTestFloat16")
+{
+    ElementwiseBinarySimple3DEndToEnd<armnn::DataType::Float16>(gpuFsaDefaultBackends, BinaryOperation::Add);
+}
+
 }
diff --git a/src/backends/gpuFsa/test/GpuFsaLayerSupportTests.cpp b/src/backends/gpuFsa/test/GpuFsaLayerSupportTests.cpp
index f162df0..49ddade 100644
--- a/src/backends/gpuFsa/test/GpuFsaLayerSupportTests.cpp
+++ b/src/backends/gpuFsa/test/GpuFsaLayerSupportTests.cpp
@@ -1,5 +1,5 @@
 //
-// Copyright © 2022-2023 Arm Ltd and Contributors. All rights reserved.
+// Copyright © 2022-2024 Arm Ltd and Contributors. All rights reserved.
 // SPDX-License-Identifier: MIT
 //
 
@@ -61,4 +61,24 @@
     REQUIRE(reasonIfNotSupported.find("NCHW not supported by this kernel") != std::string::npos);
 }
 
+TEST_CASE("IsLayerSupportedGpuFsaElementWiseBinaryAdd")
+{
+    TensorInfo input0Info({ 2, 2 }, DataType::Float32);
+    TensorInfo input1Info({ 2, 2 }, DataType::Float32);
+    TensorInfo outputInfo({ 2, 2 }, DataType::Float32);
+
+    ElementwiseBinaryDescriptor desc;
+    desc.m_Operation = BinaryOperation::Add;
+
+    GpuFsaLayerSupport supportChecker;
+    std::string reasonIfNotSupported;
+    auto supported = supportChecker.IsLayerSupported(LayerType::ElementwiseBinary,
+                                                     {input0Info, input1Info, outputInfo},
+                                                     desc,
+                                                     EmptyOptional(),
+                                                     EmptyOptional(),
+                                                     reasonIfNotSupported);
+    CHECK(supported);
+}
+
 }
\ No newline at end of file
diff --git a/src/backends/gpuFsa/test/GpuFsaOptimizedNetworkTests.cpp b/src/backends/gpuFsa/test/GpuFsaOptimizedNetworkTests.cpp
index 7e094ce..673a52a 100644
--- a/src/backends/gpuFsa/test/GpuFsaOptimizedNetworkTests.cpp
+++ b/src/backends/gpuFsa/test/GpuFsaOptimizedNetworkTests.cpp
@@ -1,5 +1,5 @@
 //
-// Copyright © 2022-2023 Arm Ltd and Contributors. All rights reserved.
+// Copyright © 2022-2024 Arm Ltd and Contributors. All rights reserved.
 // SPDX-License-Identifier: MIT
 //
 
@@ -134,4 +134,52 @@
                         &IsLayerOfType<OutputLayer>));
 }
 
+TEST_CASE("ElementwiseBinaryAddSupportedOptimizedNetwork")
+{
+    using namespace armnn;
+
+    const float qScale = 1.0f;
+    const int32_t qOffset = 0;
+
+    const TensorShape& input1Shape  = { 2, 2, 2 };
+    const TensorShape& input2Shape  = { 2, 2, 2 };
+    const TensorShape& outputShape  = { 2, 2, 2 };
+
+    TensorInfo input1TensorInfo(input1Shape, DataType::Float32, qScale, qOffset, true);
+    TensorInfo input2TensorInfo(input2Shape, DataType::Float32, qScale, qOffset, true);
+    TensorInfo outputTensorInfo(outputShape, DataType::Float32, qScale, qOffset);
+
+    IRuntime::CreationOptions options;
+    IRuntimePtr runtime(IRuntime::Create(options));
+    INetworkPtr network(INetwork::Create());
+
+    IConnectableLayer* input1 = network->AddInputLayer(0, "input0");
+    IConnectableLayer* input2 = network->AddInputLayer(1, "input1");
+
+    ElementwiseBinaryDescriptor desc;
+    desc.m_Operation = BinaryOperation::Add;
+
+    IConnectableLayer* elementwiseBinaryLayer = network->AddElementwiseBinaryLayer(desc, "elementwiseBinary");
+    IConnectableLayer* output = network->AddOutputLayer(2, "output");
+
+    Connect(input1, elementwiseBinaryLayer, input1TensorInfo, 0, 0);
+    Connect(input2, elementwiseBinaryLayer, input2TensorInfo, 0, 1);
+    Connect(elementwiseBinaryLayer, output, outputTensorInfo, 0, 0);
+
+    std::vector<BackendId> backends = { "GpuFsa" };
+
+    OptimizerOptionsOpaque optimizedOptions;
+    IOptimizedNetworkPtr optNet = Optimize(*network, backends, runtime->GetDeviceSpec(), optimizedOptions);
+    CHECK(optNet);
+
+    Graph& graph = GetGraphForTesting(optNet.get());
+
+    // Check graph layer sequence to ensure that the network has been replaced with a PreCompiledLayer
+    CHECK(CheckSequence(graph.cbegin(), graph.cend(),
+                        &IsLayerOfType<InputLayer>,
+                        &IsLayerOfType<InputLayer>,
+                        &IsLayerOfType<PreCompiledLayer>,
+                        &IsLayerOfType<OutputLayer>));
+}
+
 }
\ No newline at end of file