Spatio-temporal variation of reference evaporation (ET_0) in Shanxi Province was calculated using the Penman-Monteith model based on daily climatic data collected from 35 meteorological stations in Shanxi and its proximal regions from 1957 to 2014, using a simple linear regression with an inverse distance-weighting method. We also analyzed the factors that affect the variation of ET_0 using a stepwise regression. The results showed that:①The change of annual average ET_0 fluctuated but with a declining trend.②The multi-year average of ET_0 varied spatially with two peaks, two sub-peaks and two troughs; the spatial distribution of multi-yearaverage of ET_0 was similar to the distribution of ET_0 in autumn, but differed considerably from that in spring, winter and summer.③ In each meteorological station, the factors that significantly affected the annual average ET_0 was in order of wind speed, temperature, relative humidity, sunshine hours or precipitation, while for the average over the province, the factors that affected the ET_0 significantly was in the order of wind speed, temperature, relative humidity, sunshine hours and precipitation. Also, the impact of altitude on ET_0 was more significant than latitude.