diff --git a/python/paddle/__init__.py b/python/paddle/__init__.py
index 6f1929c64275b0..6f883b761eb326 100644
--- a/python/paddle/__init__.py
+++ b/python/paddle/__init__.py
@@ -457,6 +457,7 @@
     scale,
     sgn,
     sign,
+    signbit,
     sin,
     sin_,
     sinh,
@@ -919,4 +920,5 @@
     "index_fill_",
     'diagonal_scatter',
     'combinations',
+    'signbit',
 ]
diff --git a/python/paddle/tensor/__init__.py b/python/paddle/tensor/__init__.py
index 4c453492f193d8..160e4b26180cd6 100644
--- a/python/paddle/tensor/__init__.py
+++ b/python/paddle/tensor/__init__.py
@@ -768,6 +768,7 @@
     'atleast_3d',
     'diagonal_scatter',
     "combinations",
+    'signbit',
 ]
 
 # this list used in math_op_patch.py for magic_method bind
diff --git a/python/paddle/tensor/math.py b/python/paddle/tensor/math.py
index 688b08c597ce01..f98e2b29745762 100644
--- a/python/paddle/tensor/math.py
+++ b/python/paddle/tensor/math.py
@@ -4632,6 +4632,7 @@ def sign(x, name=None):
                 'int32',
                 'int64',
                 'float16',
+                'bfloat16',
                 'float32',
                 'float64',
             ],
@@ -7169,3 +7170,62 @@ def combinations(x, r=2, with_replacement=False, name=None):
         grids[i] = grids[i].masked_select(mask)
 
     return paddle.stack(grids, 1)
+
+
+def signbit(x, name=None):
+    r"""
+    Tests if each element of input has its sign bit set or not.
+
+    Args:
+        x (Tensor): The input Tensor. Must be one of the following types: float16, float32, float64, bfloat16, uint8, int8, int16, int32, int64.
+        name (str, optional): Name for the operation (optional, default is None).For more information, please refer to :ref:`api_guide_Name`.
+
+    Returns:
+        out (Tensor): The output Tensor. The sign bit of the corresponding element of the input tensor, True means negative, False means positive.
+
+    Examples:
+        .. code-block:: python
+            :name: signbit-example-1
+
+            >>> import paddle
+            >>> paddle.set_device('cpu')
+            >>> x = paddle.to_tensor([-0., 1.1, -2.1, 0., 2.5], dtype='float32')
+            >>> res = paddle.signbit(x)
+            >>> print(res)
+            Tensor(shape=[5], dtype=bool, place=Place(cpu), stop_gradient=True,
+            [True, False, True, False, False])
+
+        .. code-block:: python
+            :name: signbit-example-2
+
+            >>> import paddle
+            >>> paddle.set_device('cpu')
+            >>> x = paddle.to_tensor([-5, -2, 3], dtype='int32')
+            >>> res = paddle.signbit(x)
+            >>> print(res)
+            Tensor(shape=[3], dtype=bool, place=Place(cpu), stop_gradient=True,
+            [True , True , False])
+    """
+    if not isinstance(x, (paddle.Tensor, Variable)):
+        raise TypeError(f"x must be tensor type, but got {type(x)}")
+
+    check_variable_and_dtype(
+        x,
+        "x",
+        [
+            'float16',
+            'float32',
+            'float64',
+            'bfloat16',
+            'uint8',
+            'int8',
+            'int16',
+            'int32',
+            'int64',
+        ],
+        "signbit",
+    )
+    neg_zero_x = paddle.to_tensor(np.copysign(1, x.numpy()), dtype=x.dtype)
+    x = paddle.sign(neg_zero_x)
+    out = paddle.cast(x < 0, dtype='bool')
+    return out
diff --git a/test/legacy_test/test_signbit.py b/test/legacy_test/test_signbit.py
new file mode 100644
index 00000000000000..7969689acfea5c
--- /dev/null
+++ b/test/legacy_test/test_signbit.py
@@ -0,0 +1,96 @@
+# Copyright (c) 2023 PaddlePaddle Authors. All Rights Reserved.
+#
+# 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.
+
+import unittest
+
+import numpy as np
+
+import paddle
+from paddle.base import core
+
+
+def ref_np_signbit(x: np.ndarray):
+    return np.signbit(x)
+
+
+class TestSignbitAPI(unittest.TestCase):
+    def setUp(self) -> None:
+        self.cuda_support_dtypes = [
+            'float32',
+            'float64',
+            'uint8',
+            'int8',
+            'int16',
+            'int32',
+            'int64',
+        ]
+        self.cpu_support_dtypes = [
+            'float32',
+            'float64',
+            'uint8',
+            'int8',
+            'int16',
+            'int32',
+            'int64',
+        ]
+        self.place = [paddle.CPUPlace()]
+        if core.is_compiled_with_cuda():
+            self.place.append(paddle.CUDAPlace(0))
+
+    def test_dtype(self):
+        def run(place):
+            paddle.disable_static(place)
+            if core.is_compiled_with_cuda():
+                support_dtypes = self.cuda_support_dtypes
+            else:
+                support_dtypes = self.cpu_support_dtypes
+
+            for dtype in support_dtypes:
+                x = paddle.to_tensor(
+                    np.random.randint(-10, 10, size=[12, 20, 2]).astype(dtype)
+                )
+                paddle.signbit(x)
+
+        for place in self.place:
+            run(place)
+
+    def test_float(self):
+        def run(place):
+            paddle.disable_static(place)
+            if core.is_compiled_with_cuda():
+                support_dtypes = self.cuda_support_dtypes
+            else:
+                support_dtypes = self.cpu_support_dtypes
+
+            for dtype in support_dtypes:
+                np_x = np.random.randint(-10, 10, size=[12, 20, 2]).astype(
+                    dtype
+                )
+                x = paddle.to_tensor(np_x)
+                out = paddle.signbit(x)
+                np_out = out.numpy()
+                out_expected = ref_np_signbit(np_x)
+                np.testing.assert_allclose(np_out, out_expected, rtol=1e-05)
+
+        for place in self.place:
+            run(place)
+
+    def test_input_type(self):
+        with self.assertRaises(TypeError):
+            x = np.random.randint(-10, 10, size=[12, 20, 2]).astype('float32')
+            x = paddle.signbit(x)
+
+
+if __name__ == "__main__":
+    unittest.main()