This　paper　investigates　the　potential　of　ENVISAT　ASAR　data　for　agriculture　crop　mapping.　The　multi-temporal　ASAR　data　with　different　polarization　are　used.　The　study　area　is　located　in　Zhangzhou,　Fujian　province,　China.　Based　on　the　analysis　of　radar　signature　of　different　land　cover,　crop　types　like　rice,　water　bamboo　and　banana　are　classified　using　object　oriented　classifier.　Compared　to　the　field　survey　data,　the　accuracy　of　rice,　banana　and　water　bamboo　in　study　area　is　more　than　80%.　The　object　oriented　classifier　is　suitable　for　SAR　data　classification.