diff --git a/python/paddle/v2/fluid/tests/book/test_recognize_digits.py b/python/paddle/v2/fluid/tests/book/test_recognize_digits.py
index 4ecdcdc63275e33d358e3aca90520c137d964c78..19b93a66f4799ffae069e2a8be66931daef6d968 100644
--- a/python/paddle/v2/fluid/tests/book/test_recognize_digits.py
+++ b/python/paddle/v2/fluid/tests/book/test_recognize_digits.py
@@ -16,6 +16,7 @@ import argparse
 import paddle.v2.fluid as fluid
 import paddle.v2 as paddle
 import sys
+import numpy
 
 
 def parse_arg():
@@ -100,6 +101,8 @@ def main():
     else:
         avg_loss, acc = net_conf(img, label)
 
+    test_program = fluid.default_main_program().clone()
+
     optimizer = fluid.optimizer.Adam(learning_rate=0.001)
     optimizer.minimize(avg_loss)
 
@@ -112,6 +115,8 @@ def main():
         paddle.reader.shuffle(
             paddle.dataset.mnist.train(), buf_size=500),
         batch_size=BATCH_SIZE)
+    test_reader = paddle.batch(
+        paddle.dataset.mnist.test(), batch_size=BATCH_SIZE)
     feeder = fluid.DataFeeder(feed_list=[img, label], place=place)
 
     PASS_NUM = 100
@@ -119,21 +124,27 @@ def main():
         for batch_id, data in enumerate(train_reader()):
             need_check = (batch_id + 1) % 10 == 0
 
+            # train a mini-batch, fetch nothing
+            exe.run(feed=feeder.feed(data))
             if need_check:
-                fetch_list = [avg_loss, acc]
-            else:
-                fetch_list = []
-
-            outs = exe.run(feed=feeder.feed(data), fetch_list=fetch_list)
-            if need_check:
-                avg_loss_np, acc_np = outs
-                if float(acc_np) > 0.9:
+                acc_set = []
+                avg_loss_set = []
+                for test_data in test_reader():
+                    acc_np, avg_loss_np = exe.run(program=test_program,
+                                                  feed=feeder.feed(test_data),
+                                                  fetch_list=[acc, avg_loss])
+                    acc_set.append(float(acc_np))
+                    avg_loss_set.append(float(avg_loss_np))
+                # get test acc and loss
+                acc_val = numpy.array(acc_set).mean()
+                avg_loss_val = numpy.array(avg_loss_set).mean()
+                if float(acc_val) > 0.85:  # test acc > 85%
                     exit(0)
                 else:
                     print(
-                        'PassID {0:1}, BatchID {1:04}, Loss {2:2.2}, Acc {3:2.2}'.
+                        'PassID {0:1}, BatchID {1:04}, Test Loss {2:2.2}, Acc {3:2.2}'.
                         format(pass_id, batch_id + 1,
-                               float(avg_loss_np), float(acc_np)))
+                               float(avg_loss_val), float(acc_val)))
 
 
 if __name__ == '__main__':