Daily maximum and minimum temperature data of 34 meteorological stations in Inner Mongolia during 1951 - 2014 were selected in this study. Multiple methods,including linear trend,anomaly,accumulative anomaly, Mann - Kendall test and spatial analysis,were applied to detect spatial and temporal characteristics of extreme temperature events in Inner Mongolia. The results showed that (1) frost days,ice days and cool days presented a significantly decreasing trend,and the trend magnitudes were 3. 3d /10a,1. 5d /10a and 1. 6d /10a, respectively. Warm days showed a significantly increasing trend,and the rate was 1. 7d /10a. (2) Taking the middle of 1980s as a boundary,anomaly and accumulative anomaly of frost days,ice days and cool days could be divided into two steps. The annual values were higher than the long - term average during most years of the former stage,and the accumulative anomaly exhibited an increased trend. However,the later stage had inverse changes. Takingthe middle of 1990s as a boundary,the anomaly and accumulative anomaly of warm days had contrary variations with the frost days,ice days and cool days. It indicated that the indices of extreme low temperature events (frost days,ice days and cool days) were earlier than extreme high temperature events (warm days) in answer to global warming. (3) There was a sudden changes of frost days,ice days and cool days in 1988,1992 and 1991,respectively. (4) Frost days and ice days have increased from low latitude (altitude) to high latitude (altitude) in Inner Mongolia,which illustrated the rules of regional differentiation. (5) The higher value of cool days and warm days concentrated on the middle of Inner Mongolia. (6) Frost days,ice days and cool days of every area in Inner Mongolia had a decreasing trend,and warm days had an increasing trend,which was consistent with the temporal result. Generally the most significant trends in the frost days and ice days could be found at the area with higher intensity of human activity,indicating temperature changes were possibly affected by human activity to some extent.