diff --git a/inference-engine/samples/speech_sample/main.cpp b/inference-engine/samples/speech_sample/main.cpp
index 2ae1be9e0e3..ef36d91732f 100644
--- a/inference-engine/samples/speech_sample/main.cpp
+++ b/inference-engine/samples/speech_sample/main.cpp
@@ -602,41 +602,50 @@ int main(int argc, char *argv[]) {
         }
 
         if (FLAGS_q.compare("user") == 0) {
-            auto scaleFactorInput = ParseScaleFactors(FLAGS_sf);
-            if (numInputArkFiles != scaleFactorInput.size()) {
-                std::string errMessage("Incorrect command line for multiple inputs: "
-                    + std::to_string(scaleFactorInput.size()) + " scale factors provided for "
-                    + std::to_string(numInputArkFiles) + " input files.");
-                throw std::logic_error(errMessage);
-            }
+            if (!FLAGS_rg.empty()) {
+                slog::warn << "Custom scale factor will be ignored - using scale factor from provided imported gna model: "
+                           << FLAGS_rg << slog::endl;
+            } else {
+                auto scaleFactorInput = ParseScaleFactors(FLAGS_sf);
+                if (numInputArkFiles != scaleFactorInput.size()) {
+                    std::string errMessage("Incorrect command line for multiple inputs: "
+                        + std::to_string(scaleFactorInput.size()) + " scale factors provided for "
+                        + std::to_string(numInputArkFiles) + " input files.");
+                    throw std::logic_error(errMessage);
+                }
 
-            for (size_t i = 0; i < scaleFactorInput.size(); ++i) {
-                slog::info << "For input " << i << " using scale factor of " << scaleFactorInput[i] << slog::endl;
-                std::string scaleFactorConfigKey = GNA_CONFIG_KEY(SCALE_FACTOR) + std::string("_") + std::to_string(i);
-                gnaPluginConfig[scaleFactorConfigKey] = scaleFactorInput[i];
+                for (size_t i = 0; i < scaleFactorInput.size(); ++i) {
+                    slog::info << "For input " << i << " using scale factor of " << scaleFactorInput[i] << slog::endl;
+                    std::string scaleFactorConfigKey = GNA_CONFIG_KEY(SCALE_FACTOR) + std::string("_") + std::to_string(i);
+                    gnaPluginConfig[scaleFactorConfigKey] = scaleFactorInput[i];
+                }
             }
         } else {
             // "static" quantization with calculated scale factor
-            for (size_t i = 0; i < numInputArkFiles; i++) {
-                auto inputArkName = inputArkFiles[i].c_str();
-                std::string name;
-                std::vector<uint8_t> ptrFeatures;
-                uint32_t numArrays(0), numBytes(0), numFrames(0), numFrameElements(0), numBytesPerElement(0);
-                GetKaldiArkInfo(inputArkName, 0, &numArrays, &numBytes);
-                ptrFeatures.resize(numBytes);
-                LoadKaldiArkArray(inputArkName,
-                                  0,
-                                  name,
-                                  ptrFeatures,
-                                  &numFrames,
-                                  &numFrameElements,
-                                  &numBytesPerElement);
-                auto floatScaleFactor =
+            if (!FLAGS_rg.empty()) {
+                slog::info << "Using scale factor from provided imported gna model: " << FLAGS_rg << slog::endl;
+            } else {
+                for (size_t i = 0; i < numInputArkFiles; i++) {
+                    auto inputArkName = inputArkFiles[i].c_str();
+                    std::string name;
+                    std::vector<uint8_t> ptrFeatures;
+                    uint32_t numArrays(0), numBytes(0), numFrames(0), numFrameElements(0), numBytesPerElement(0);
+                    GetKaldiArkInfo(inputArkName, 0, &numArrays, &numBytes);
+                    ptrFeatures.resize(numBytes);
+                    LoadKaldiArkArray(inputArkName,
+                        0,
+                        name,
+                        ptrFeatures,
+                        &numFrames,
+                        &numFrameElements,
+                        &numBytesPerElement);
+                    auto floatScaleFactor =
                         ScaleFactorForQuantization(ptrFeatures.data(), MAX_VAL_2B_FEAT, numFrames * numFrameElements);
-                slog::info << "Using scale factor of " << floatScaleFactor << " calculated from first utterance."
-                           << slog::endl;
-                std::string scaleFactorConfigKey = GNA_CONFIG_KEY(SCALE_FACTOR) + std::string("_") + std::to_string(i);
-                gnaPluginConfig[scaleFactorConfigKey] = std::to_string(floatScaleFactor);
+                    slog::info << "Using scale factor of " << floatScaleFactor << " calculated from first utterance."
+                        << slog::endl;
+                    std::string scaleFactorConfigKey = GNA_CONFIG_KEY(SCALE_FACTOR) + std::string("_") + std::to_string(i);
+                    gnaPluginConfig[scaleFactorConfigKey] = std::to_string(floatScaleFactor);
+                }
             }
         }
 
diff --git a/inference-engine/src/gna_plugin/gna_plugin.cpp b/inference-engine/src/gna_plugin/gna_plugin.cpp
index 92b6a695eea..65bdde36d28 100644
--- a/inference-engine/src/gna_plugin/gna_plugin.cpp
+++ b/inference-engine/src/gna_plugin/gna_plugin.cpp
@@ -1180,6 +1180,12 @@ InferenceEngine::IExecutableNetwork::Ptr GNAPlugin::ImportNetwork(std::istream&
     auto serial = GNAModelSerial(&std::get<0>(nnets.back())->obj, mt);
 #endif
 
+    if (!inputsDesc->inputScaleFactors.empty()) {
+        gnalog() << "[Import Network] Clearing input scale factors defined in configuration. "
+                 << "Scale factors provided in imported model will be used\n";
+        inputsDesc->inputScaleFactors.clear();
+    }
+
     serial.setHeader(header);
     serial.Import(basePtr,
             header.gnaMemSize,
diff --git a/inference-engine/tests/functional/plugin/gna/import_export_network.cpp b/inference-engine/tests/functional/plugin/gna/import_export_network.cpp
index 663113c8381..9dbe1b2a78a 100644
--- a/inference-engine/tests/functional/plugin/gna/import_export_network.cpp
+++ b/inference-engine/tests/functional/plugin/gna/import_export_network.cpp
@@ -23,7 +23,8 @@
 typedef std::tuple<
     InferenceEngine::Precision,         // Network Precision
     std::string,                        // Target Device
-    std::map<std::string, std::string>  //Configuration
+    std::map<std::string, std::string>, // Export Configuration
+    std::map<std::string, std::string>  // Import Configuration
 > exportImportNetworkParams;
 
 namespace LayerTestsDefinitions {
@@ -34,14 +35,18 @@ class ImportNetworkTest : public testing::WithParamInterface<exportImportNetwork
         static std::string getTestCaseName(testing::TestParamInfo<exportImportNetworkParams> obj) {
             InferenceEngine::Precision netPrecision;
             std::string targetDevice;
-            std::map<std::string, std::string> configuration;
-            std::tie(netPrecision, targetDevice, configuration) = obj.param;
+            std::map<std::string, std::string> exportConfiguration;
+            std::map<std::string, std::string> importConfiguration;
+            std::tie(netPrecision, targetDevice, exportConfiguration, importConfiguration) = obj.param;
 
             std::ostringstream result;
             result << "netPRC=" << netPrecision.name() << "_";
             result << "targetDevice=" << targetDevice << "_";
-            for (auto const& configItem : configuration) {
-                result << "_configItem=" << configItem.first << "_" << configItem.second;
+            for (auto const& configItem : exportConfiguration) {
+                result << "_exportConfigItem=" << configItem.first << "_" << configItem.second;
+            }
+            for (auto const& configItem : importConfiguration) {
+                result << "_importConfigItem=" << configItem.first << "_" << configItem.second;
             }
             return result.str();
         }
@@ -49,6 +54,7 @@ class ImportNetworkTest : public testing::WithParamInterface<exportImportNetwork
         void Run() override {
             SKIP_IF_CURRENT_TEST_IS_DISABLED()
 
+            configuration.insert(exportConfiguration.begin(), exportConfiguration.end());
             LoadNetwork();
             Infer();
             executableNetwork.Export("exported_model.blob");
@@ -57,6 +63,9 @@ class ImportNetworkTest : public testing::WithParamInterface<exportImportNetwork
             auto referenceOutputs = CalculateRefs();
             Compare(referenceOutputs, actualOutputs);
 
+            for (auto const& configItem : importConfiguration) {
+                configuration[configItem.first] = configItem.second;
+            }
             std::fstream inputStream("exported_model.blob", std::ios_base::in | std::ios_base::binary);
             if (inputStream.fail()) {
                 FAIL() << "Cannot open file to import model: exported_model.blob";
@@ -68,7 +77,7 @@ class ImportNetworkTest : public testing::WithParamInterface<exportImportNetwork
     protected:
         void SetUp() override {
             InferenceEngine::Precision netPrecision;
-            std::tie(netPrecision, targetDevice, configuration) = this->GetParam();
+            std::tie(netPrecision, targetDevice, exportConfiguration, importConfiguration) = this->GetParam();
             auto ngPrc = FuncTestUtils::PrecisionUtils::convertIE2nGraphPrc(netPrecision);
 
             auto params = ngraph::builder::makeParams(ngPrc, { {1, 336} });
@@ -95,6 +104,9 @@ class ImportNetworkTest : public testing::WithParamInterface<exportImportNetwork
         }
 
     private:
+        std::map<std::string, std::string> exportConfiguration;
+        std::map<std::string, std::string> importConfiguration;
+
         std::vector<std::vector<std::uint8_t>> CalculateImportedNetwork(std::istream& networkModel) {
             auto importedNetwork = core->ImportNetwork(networkModel, targetDevice, configuration);
 
@@ -149,18 +161,30 @@ class ImportNetworkTest : public testing::WithParamInterface<exportImportNetwork
         InferenceEngine::Precision::FP16
     };
 
-    const std::vector<std::map<std::string, std::string>> configs = {
+    const std::vector<std::map<std::string, std::string>> exportConfigs = {
         {
             {"GNA_DEVICE_MODE", "GNA_SW_EXACT"},
             {"GNA_SCALE_FACTOR_0", "327.67"}
         }
     };
 
+    const std::vector<std::map<std::string, std::string>> importConfigs = {
+        {
+            {"GNA_DEVICE_MODE", "GNA_SW_EXACT"},
+            {"GNA_SCALE_FACTOR_0", "32767"}
+        },
+        {
+            {"GNA_DEVICE_MODE", "GNA_SW_EXACT"},
+            {"GNA_SCALE_FACTOR_0", "327.67"}
+        },
+    };
+
     INSTANTIATE_TEST_CASE_P(ImportNetworkCase, ImportNetworkTest,
         ::testing::Combine(
             ::testing::ValuesIn(netPrecisions),
             ::testing::Values(CommonTestUtils::DEVICE_GNA),
-            ::testing::ValuesIn(configs)),
+            ::testing::ValuesIn(exportConfigs),
+            ::testing::ValuesIn(importConfigs)),
         ImportNetworkTest::getTestCaseName);
 
 } // namespace LayerTestsDefinitions