Abstract: Baiyin District of Baiyin City，Gansu Province，China relies on agriculture and stockbreeding enterprises to develop featured agriculture. The action can promote agricultural development which makes farmers’ incomes growing. However，in fact，Baiyin District is located in a resource-based city. Industrial and mineral industries include mining，metal smelting and electroplating are relatively developed. Cultivated land has been polluted to a certain extent which affects the health of citizens. So，it is necessary to understand the status quo of the cultivated land pollution. At present，the heavy metal pollution research is mainly focused on the mechanism，prevention and cure，repair，pollution status quo and risk assessment，etc. Most studies use less data to do risk assessment that can not comprehensively analyze the situation of heavy metal pollutions. Therefore，the paper gathered 1 070 sampling points on arable layer soil and detect 8 kinds of heavy metals of it to enrich the data. Then，via GIS，the paper evaluated the pollution present situation of Baiyin District，providing a basis for heavy metal pollution control and agriculture distribution. Heavy metal contents of Cu，Zn，Ni，Pb，As，Cr，V，Co were measured with handheld X-ray fluorescence spectrometry （X-MET7000）. And the spatial distribution data of heavy metals were obtained with Geostatistical analyst of ArcGIS9.3 and SPSS18.0. Meanwhile，combining with the newest soil environment quality standard （GB15618- 2008） and using single factor and Nemerow index method, comprehensive pollution index method，the analysis was carried out on soil heavy metal spatial distribution and pollution levels of cultivated land arable layer. Conclusions include the following three aspects：descriptive statistics，spatial variation characters and spatial distribution. Descriptive statistics shows that the average pollution degrees of the 8 heavy metals are ranked from big to small order as follows：As，V，Pb，Zn，Cu，Co，Ni，Cr，while the variation degrees of this 8 heavy metals as follows：Pb，Cu，Zn，Co，As，Cr，V，Ni. Spatial variation characteristics of soil shows that the optimal fitting model for Cu，Pb is Pentaspherical Model；for Zn，V is Gaussion Model；for Ni，As is Exponential Model；for Cr，Co is Circular Model. And the spatial correlation from big to small order is Zn，Cu，As，Pb，Ni，Cr，V，Co. Spatial distribution maps of soil heavy metal content show that Cu，Zn，Pb and As have obvious zonality distributions and distribute like islands，while Ni，Cr，V and Co have interlaced distributions. Finally，through the analysis of the pollution status and pollution evaluation，the paper found that Baiyin District was contaminated by Cu，Zn，Pb and As. Contaminated farmland accounts for 6.56% of the total area，covers an area of 763.4 hm²，the alert level farmland may be polluted by daily human activities and contaminated farmland may be caused by heavy metal mining and sewage irrigation．

Key words: GIS, Baiyin District, arable layer soil, heavy metals, spatial variability, pollution evaluation