diff --git a/CMakeLists.txt b/CMakeLists.txt
index 265ddc9504167f21f54a1b1e7777147b3b6d37d9..586628dc48d2d61d130fb2719c45150fdb41a01e 100755
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -107,6 +107,7 @@ option(WITH_LITE   "Compile Paddle Fluid with Lite Engine" OFF)
 option(WITH_NCCL   "Compile PaddlePaddle with NCCL support"             ON)
 option(WITH_CRYPTO   "Compile PaddlePaddle with crypto support"         ON)
 option(WITH_ARM   "Compile PaddlePaddle with arm support"         OFF)
+option(WITH_NVINFER_PLUGIN   "Use plugins of nvinfer_plugin"         OFF)
 
 # PY_VERSION
 if(NOT PY_VERSION)
@@ -234,6 +235,9 @@ if(ON_INFER)
     message(STATUS "On inference mode, will take place some specific optimization.")
     include(inference_lib)
     add_definitions(-DPADDLE_ON_INFERENCE)
+    if (WITH_NVINFER_PLUGIN)
+        add_definitions(-DUSE_NVINFER_PLUGIN)
+    endif()
 else()
     #TODO(luotao), combine this warning with `make inference_lib_dist` command.
     message(WARNING "On inference mode, will take place some specific optimization. Turn on the ON_INFER flag when building inference_lib only.")