diff --git a/python/paddle/fluid/tests/unittests/test_sparse_reshape_op.py b/python/paddle/fluid/tests/unittests/test_sparse_reshape_op.py
index 34b7a95299f3e05fb35157e0849f3f8732fb8ab9..84315ede556ac22b47b3708605a488e063c5ff32 100644
--- a/python/paddle/fluid/tests/unittests/test_sparse_reshape_op.py
+++ b/python/paddle/fluid/tests/unittests/test_sparse_reshape_op.py
@@ -36,6 +36,8 @@ class TestReshape(unittest.TestCase):
         paddle.sparse.reshape.
         """
         mask = np.random.randint(0, 2, x_shape)
+        while np.sum(mask) == 0:
+            mask = paddle.randint(0, 2, x_shape)
         np_x = np.random.randint(-100, 100, x_shape) * mask
 
         # check cpu kernel
diff --git a/python/paddle/fluid/tests/unittests/test_sparse_transpose_op.py b/python/paddle/fluid/tests/unittests/test_sparse_transpose_op.py
index 32772d389138300786b2057a184b21690f7704a6..74a9d7e3308b3d3089871fc66b3f4b17f804c33c 100644
--- a/python/paddle/fluid/tests/unittests/test_sparse_transpose_op.py
+++ b/python/paddle/fluid/tests/unittests/test_sparse_transpose_op.py
@@ -23,6 +23,8 @@ class TestTranspose(unittest.TestCase):
     # x: sparse, out: sparse
     def check_result(self, x_shape, dims, format):
         mask = paddle.randint(0, 2, x_shape).astype("float32")
+        while paddle.sum(mask) == 0:
+            mask = paddle.randint(0, 2, x_shape).astype("float32")
         # "+ 1" to make sure that all zero elements in "origin_x" is caused by multiplying by "mask",
         # or the backward checks may fail.
         origin_x = (paddle.rand(x_shape, dtype='float32') + 1) * mask
diff --git a/python/paddle/fluid/tests/unittests/test_sparse_unary_op.py b/python/paddle/fluid/tests/unittests/test_sparse_unary_op.py
index edb7393bf305631e25af4d0dff8eaaafcc676c65..908121ace01181faf339f2f618f3d1fc4ccde86b 100644
--- a/python/paddle/fluid/tests/unittests/test_sparse_unary_op.py
+++ b/python/paddle/fluid/tests/unittests/test_sparse_unary_op.py
@@ -30,6 +30,8 @@ class TestSparseUnary(unittest.TestCase):
     def check_result(self, dense_func, sparse_func, format, *args):
         origin_x = paddle.rand([8, 16, 32], dtype='float32')
         mask = paddle.randint(0, 2, [8, 16, 32]).astype('float32')
+        while paddle.sum(mask) == 0:
+            mask = paddle.randint(0, 2, [8, 16, 32]).astype("float32")
 
         # --- check sparse coo with dense --- #
         dense_x = origin_x * mask