This paper constructs the evaluation index system of farmland conversion pressure in Hubei province from economic factors,social factors and resource endowments,and uses SOM (Self-organizing feature Map) to gauge the pressure of farmland conversion in 16 cities of Hubei province from 2006 to 2014.The result shows that firstly,there are some regional differences of the farmland conversion pressure in 16 cities of Hubei province.In 2006,this pressure showed decreasing trend from southwest,northeast to the central regions of this province.In 2008,there was little change in regional differences.In 2010,it showed the marginalized pattern.In 2012，it showed that Wuhan city,Yichang city are the centers,the pressure was decreasing to the surrounding cities.Until 2014,the pressure on farmland conversion in Wuhan tends to be gentle and some small cities around Wuhan have a slight rise.Secondly,the SOM neural network model with a strong ability of clustering and faulttolerant can be used to classify farmland conversion pressure.The results reflect the spatial heterogeneity of the farmland conversion pressure in 16 cities of Hubei province,which provides a scientific reference for the optimal allocation of land resources.