diff --git a/python/paddle/fluid/dygraph/dygraph_to_static/convert_operators.py b/python/paddle/fluid/dygraph/dygraph_to_static/convert_operators.py
index e0b46fe2341a3015e2703da45bc45cb831d000e5..b78b5957393a3ee1c594014ea1a9479ab64dec12 100644
--- a/python/paddle/fluid/dygraph/dygraph_to_static/convert_operators.py
+++ b/python/paddle/fluid/dygraph/dygraph_to_static/convert_operators.py
@@ -414,6 +414,8 @@ def convert_len(var):
             # Note: Length of var may be known ahead of time in dygraph,
             # but it probably represents batch size which can be variant.
             # so we return a variable dynamically inferred from var.shape.
+            if var.shape[0] > 0 and var.type == core.VarDesc.VarType.LOD_TENSOR:
+                return var.shape[0]
             return nn.shape(var)[0]
         elif var.type == core.VarDesc.VarType.LOD_TENSOR_ARRAY:
             return control_flow.array_length(var)
@@ -447,7 +449,10 @@ class VariableTuple:
     def __init__(self, var, start=0):
         self.var = var
         self.len = convert_len(var)
-        self.rag = paddle_range(start, start + self.len, 1, paddle.int64)
+        if isinstance(self.len, Variable):
+            self.rag = paddle_range(start, start + self.len, 1, paddle.int64)
+        else:
+            self.rag = range(start, start + self.len)
 
     def __getitem__(self, idx):
         return self.rag[idx], self.var[idx]
diff --git a/python/paddle/fluid/tests/unittests/dygraph_to_static/test_len.py b/python/paddle/fluid/tests/unittests/dygraph_to_static/test_len.py
index 28f79b57b6ba34fa962bbb94155eb170ec276c31..386c3a1bd7bbe089b0a258755f2417c9a7873b76 100644
--- a/python/paddle/fluid/tests/unittests/dygraph_to_static/test_len.py
+++ b/python/paddle/fluid/tests/unittests/dygraph_to_static/test_len.py
@@ -17,12 +17,14 @@ from __future__ import print_function
 import unittest
 
 import numpy as np
+import paddle
 import paddle.fluid as fluid
 from paddle.fluid.dygraph import declarative
 from paddle.fluid.dygraph.dygraph_to_static import convert_call
 
 SEED = 2020
 np.random.seed(SEED)
+paddle.enable_static()
 
 
 def len_with_tensor(x):