Abstract: Yellow-River basin is located in northern center of China， mainly influenced by continental monsoon climate. Ice disasters take place during the period of freeze or break up with the impacted of discharge， temperature and variable river bed. Ice forecasting is critical facet to hydrology forecasting. In recent years， ice features in Inner Mongolia reach of Yellow-River have remarkably changed due to the climate warming， river channel sedimentation and human activities， which result in higher requirements for ice forecast. Therefore， ice forecast system has to be continually promoting. In this paper， the key technology of ice forecasting methods and forecasting models are introduced ， including medium-short term temperature forecast model， indices or empirical correlation methods， artificial neural network， hydrology and thermodynamics model， as well as river ice hydraulic model. It is suggested that ice hydraulic forecast model need to be build up based on ice condition observation and research on ice regulation in river.