Studying grassland phenology and its relation to meteorological factors is crucial for enhancing the accuracy of net primary production and carbon sequestration estimate, and identifying appropriate grazing and harvesting times. Grasslands of Leymus chinensis are an endemic and dominant formation in the eastern part of Eurasia. Many studies have focused on physiological characteristics but few on phenological performances of Leymus chinensis in Inner Mongolia. Therefore, we used phenological observation data of Leymus chinensis and corresponding meteorological data from seven stations of Inner Mongolia during 1983 to 2002 to reveal statistical relationships between the greenup/brownoff date of Leymus chinensis and mean air temperature/precipitation, and find out principal factors influencing greenup and brownoff dates of Leymus chinensis at regional and local scales. Annual mean air temperature showed a significant increase trend at all seven stations, whereas annual mean precipitation indicated an insignificant decrease trend at most of the stations. At the same time, both greenup and brownoff dates of Leymus chinensis represented a dominantly significant advancement, and the growing season (period from greenup to brownoff) was shortened at most of the stations. So, the response of Leymus chinensis growing season length to climate change was different from that of woody plants, namely, a temperature increase may not induce a growing season lengthening of Leymus chinensis. At regional scales, the most significantly negative correlation appeared between the Leymus chinensis greenup date and mean air temperature over the previous one month, namely, if the mean air temperature increased 1℃, the greenup date advanced about 2.4 days. At local scales, Leymus chinensis greenup at stations of the forest steppe area was influenced mainly by the mean air temperature over the previous one month (negative correlation), whereas Leymus chinensis greenup at stations of the typical steppe area was influenced primarily by the cumulative precipitation from the last October or November to the greenup date (negative correlation). A special case occurred at Eerguna station located in the most northern part of the forest steppe area, there Leymus chinensis greenup was affected by both mean air temperature over the previous one month and cumulative precipitation from the last October to the greenup date. In terms of relationships between the Leymus chinensis brownoff date and meteorological factors, at regional scales, the most significantly negative correlation occurred between the Leymus chinensis brownoff date and mean air temperature over the previous one month, namely, if the mean air temperature increased 1℃, the brownoff date advanced about 3.7 days. By contrast, the correlation between the Leymus chinensis brownoff date and precipitation was less significant. At local scales, a negative correlation dominated between the brownoff date and the previous air temperature, whereas a positive correlation dominated between the brownoff date and the previous precipitation.