diff --git a/onnxruntime/core/optimizer/qdq_transformer/clip_quantizelinear.cc b/onnxruntime/core/optimizer/qdq_transformer/clip_quantizelinear.cc
index a1859b9d7071b..faeee1abd07fc 100644
--- a/onnxruntime/core/optimizer/qdq_transformer/clip_quantizelinear.cc
+++ b/onnxruntime/core/optimizer/qdq_transformer/clip_quantizelinear.cc
@@ -81,7 +81,7 @@ static bool GetQConstantLowerUpper(const Graph& graph, const Node& node, float&
   return true;
 }
 
-bool ClipQuantFusion::SatisfyCondition(const Graph& graph, const Node& node, const logging::Logger& /*logger*/) const {
+bool ClipQuantFusion::SatisfyCondition(const Graph& graph, const Node& node, const logging::Logger& logger) const {
   if (!graph_utils::IsSupportedOptypeVersionAndDomain(node, "Clip", {1, 6, 11, 12, 13}) ||
       !graph_utils::IsSupportedProvider(node, {kCpuExecutionProvider}) ||
       !optimizer_utils::CheckOutputEdges(graph, node, 1)) {
@@ -95,6 +95,10 @@ bool ClipQuantFusion::SatisfyCondition(const Graph& graph, const Node& node, con
     return false;
   }
 
+  if (!graph_utils::CanRemoveNode(graph, node, logger)) {
+    return false;
+  }
+
   return true;
 }
 
diff --git a/onnxruntime/test/optimizer/qdq_transformer_test.cc b/onnxruntime/test/optimizer/qdq_transformer_test.cc
index 1cda0c7965430..fe84f27a9105c 100644
--- a/onnxruntime/test/optimizer/qdq_transformer_test.cc
+++ b/onnxruntime/test/optimizer/qdq_transformer_test.cc
@@ -3221,6 +3221,37 @@ TEST(QDQTransformerTests, ReluQuantFusion_Level2Only) {
   test_case(TransformerLevel::Level3, 0);     // Will not fuse Relu into QuantizeLinear due to zero-point != -128
 }
 
+// Test skip removing edge when min/max come from DequantizeLinear nodes instead of initializers).
+TEST(QDQTransformerTests, ClipQuantFusion_MultipleInputEdges) {
+  auto build_test_case = [&](ModelTestBuilder& builder) {
+    // Clip's min coming from another DQ node (creating 2 input edges to Clip)
+    auto* input_arg = builder.MakeInput<uint8_t>({1, 2, 2, 2}, std::numeric_limits<uint8_t>::min(),
+                                                 std::numeric_limits<uint8_t>::max());
+    auto* data_dq = builder.MakeIntermediate();
+    builder.AddDequantizeLinearNode<uint8_t>(input_arg, 0.04f, static_cast<uint8_t>(0), data_dq);
+    auto* min_q = builder.MakeScalarInitializer<uint8_t>(0);
+    auto* min_dq = builder.MakeIntermediate();
+    builder.AddDequantizeLinearNode<uint8_t>(min_q, 0.04f, static_cast<uint8_t>(0), min_dq);
+    auto* clip_output = builder.MakeIntermediate();
+    builder.AddNode("Clip", {data_dq, min_dq}, {clip_output});
+    auto* output_q = builder.MakeIntermediate();
+    builder.AddQuantizeLinearNode<uint8_t>(clip_output, 0.04f, static_cast<uint8_t>(0), output_q);
+    auto* output_arg = builder.MakeOutput();
+    builder.AddDequantizeLinearNode<uint8_t>(output_q, 0.04f, static_cast<uint8_t>(0), output_arg);
+  };
+
+  auto check_graph = [&](InferenceSessionWrapper& session) {
+    auto op_to_count = CountOpsInGraph(session.GetGraph());
+    // ClipQuantFusion should skip it due to CanRemoveNode check
+    EXPECT_EQ(op_to_count["Clip"], 1);
+  };
+
+  TransformerTester(build_test_case, check_graph,
+                    TransformerLevel::Default,
+                    TransformerLevel::Level2,
+                    18);  // opset
+}
+
 template <typename ScaleType, typename ZpType>
 void TestWhereWithDqInput(bool is_dq_1,
                           bool is_dq_2,