Remote sensing land surface high temperature indices of Beijing are built from NOAA18/AVHRR satellite data and meteorological station data. Based on these indices, high temperature spatiotemporal patterns were analyzed using data collected from 1989 to 2008 (lack 2002). Results show that remote sensing land surface temperatures of 44, 47, and 52℃ correspond to air temperatures of 35, 37, and 40℃, respectively. In addition, the authors found that high temperature indices retrieved from remote sensing are most suitable for high temperature analysis of the plain areas of Beijing. The high temperature characteristics reflected by the land surface high temperature intensity index are about the same as those from weather station data. The land surface high temperature proportion index is presented as a useful tool for explaining the intensity and spatial differences of high temperature. In summer (from June to August) the probabilities of ten-day average land surface temperatures ≥44℃ are generally less than 50%, and are widely distributed in urban and plain area. Probabilities of ten-day average land temperatures ≥47℃ are generally less than 40%, and are concentrated within the five-circle regions. Probabilities of ten-day average land temperatures ≥52℃ are generally less than 15%, and are concentrated within urban area. Between June and September, higher probabilities of high land surface temperature are concentrated in the five-circle area. The probabilities of ten-day average land surface temperature being ≥44℃, ≥47℃, and ≥52℃ are 80%-100%, 60%-80%, and 10%-40%.