diff --git a/paddle/operators/increment_op.cc b/paddle/operators/increment_op.cc
new file mode 100644
index 0000000000000000000000000000000000000000..139392c691e00b2a94f46801f1cfc2018ce139f5
--- /dev/null
+++ b/paddle/operators/increment_op.cc
@@ -0,0 +1,75 @@
+/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+   http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License. */
+
+#include "paddle/operators/increment_op.h"
+
+namespace paddle {
+namespace operators {
+
+class IncrementOp : public framework::OperatorWithKernel {
+ public:
+  using framework::OperatorWithKernel::OperatorWithKernel;
+
+  void InferShape(framework::InferShapeContext *ctx) const override {
+    PADDLE_ENFORCE(ctx->HasInput("X"),
+                   "Input(X) of IncrementOp should not be null.");
+    PADDLE_ENFORCE(ctx->HasOutput("Out"),
+                   "Output(Out) of IncrementOp should not be null.");
+    ctx->SetOutputDim("Out", ctx->GetInputDim("X"));
+    ctx->ShareLoD("X", /*->*/ "Out");
+  }
+};
+
+template <typename AttrType>
+class IncrementOpMaker : public framework::OpProtoAndCheckerMaker {
+ public:
+  IncrementOpMaker(framework::OpProto *proto,
+                   framework::OpAttrChecker *op_checker)
+      : OpProtoAndCheckerMaker(proto, op_checker) {
+    AddInput("X", "(Tensor) The input tensor of increment operator");
+    AddOutput("Out", "(Tensor) The output tensor of increment operator.");
+    AddComment(R"DOC(Increment operator
+
+The equation is: Out = X + step
+)DOC");
+    AddAttr<AttrType>("step",
+                      "The step size by which the "
+                      "input tensor will be incremented.")
+        .SetDefault(1.0);
+  }
+};
+
+class IncrementGradOpMaker : public framework::SingleGradOpDescMaker {
+ public:
+  using framework::SingleGradOpDescMaker::SingleGradOpDescMaker;
+
+  std::unique_ptr<framework::OpDescBind> Apply() const override {
+    auto *grad_op = new framework::OpDescBind();
+    grad_op->SetType("scale");
+    grad_op->SetInput("X", OutputGrad("Out"));
+    grad_op->SetOutput("Out", InputGrad("X"));
+    grad_op->SetAttr("scale", 1.0f);
+    return std::unique_ptr<framework::OpDescBind>(grad_op);
+  }
+};
+
+}  // namespace operators
+}  // namespace paddle
+
+namespace ops = paddle::operators;
+
+REGISTER_OPERATOR(increment, ops::IncrementOp, ops::IncrementOpMaker<float>,
+                  ops::IncrementGradOpMaker);
+REGISTER_OP_CPU_KERNEL(increment,
+                       ops::IncrementKernel<paddle::platform::CPUPlace, float>);
diff --git a/paddle/operators/increment_op.cu b/paddle/operators/increment_op.cu
new file mode 100644
index 0000000000000000000000000000000000000000..659c380d147a36650452bea23b30cbcf1ff516ee
--- /dev/null
+++ b/paddle/operators/increment_op.cu
@@ -0,0 +1,19 @@
+/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+   http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License. */
+
+#include "paddle/operators/increment_op.h"
+
+REGISTER_OP_GPU_KERNEL(
+    increment,
+    paddle::operators::IncrementKernel<paddle::platform::GPUPlace, float>);
diff --git a/paddle/operators/increment_op.h b/paddle/operators/increment_op.h
new file mode 100644
index 0000000000000000000000000000000000000000..342e254fc453555c70923efbca02fdfd014af015
--- /dev/null
+++ b/paddle/operators/increment_op.h
@@ -0,0 +1,40 @@
+/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+   http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License. */
+
+#pragma once
+
+#include "paddle/framework/eigen.h"
+#include "paddle/framework/op_registry.h"
+
+namespace paddle {
+namespace operators {
+template <typename Place, typename T, typename AttrType = T>
+class IncrementKernel : public framework::OpKernel<T> {
+ public:
+  virtual void Compute(const framework::ExecutionContext& context) const {
+    auto* tensor = context.Output<framework::Tensor>("Out");
+    auto* in = context.Input<framework::Tensor>("X");
+    tensor->mutable_data<T>(in->place());
+
+    auto step = static_cast<T>(context.Attr<AttrType>("step"));
+
+    auto eigen_out = framework::EigenVector<T>::Flatten(*tensor);
+    auto eigen_in = framework::EigenVector<T>::Flatten(*in);
+    auto& place = context.GetEigenDevice<Place>();
+    eigen_out.device(place) = eigen_in + step;
+  }
+};
+
+}  // namespace operators
+}  // namespace paddle
diff --git a/python/paddle/v2/framework/tests/test_increment_op.py b/python/paddle/v2/framework/tests/test_increment_op.py
new file mode 100644
index 0000000000000000000000000000000000000000..e174272b05b9413cc2bc1e099c4dd17899829e76
--- /dev/null
+++ b/python/paddle/v2/framework/tests/test_increment_op.py
@@ -0,0 +1,41 @@
+import unittest
+import numpy as np
+from op_test import OpTest
+
+
+class TestIncrementOpPositiveStep(OpTest):
+    """Test increment op with positive step
+    """
+
+    def setUp(self):
+        self.op_type = "increment"
+        self.inputs = {'X': np.random.random((10, 10)).astype("float32")}
+        self.attrs = {'step': 14.8}
+        self.outputs = {'Out': self.inputs['X'] + self.attrs['step']}
+
+    def test_check_output(self):
+        self.check_output()
+
+    def test_check_grad(self):
+        self.check_grad(['X'], 'Out')
+
+
+class TestIncrementOpNegativeStep(OpTest):
+    """Test increment op with negative step
+    """
+
+    def setUp(self):
+        self.op_type = "increment"
+        self.inputs = {'X': np.random.random((10, 10)).astype("float32")}
+        self.attrs = {'step': -3.8}
+        self.outputs = {'Out': self.inputs['X'] + self.attrs['step']}
+
+    def test_check_output(self):
+        self.check_output()
+
+    def test_check_grad(self):
+        self.check_grad(['X'], 'Out')
+
+
+if __name__ == "__main__":
+    unittest.main()