Abstract: Abstract: In order to explore the spatial distribution and formation mechanism of the cultivated land transfer rent, the spatial agglomeration and spatial layout of the cultivated land transfer rent in Henan Province was studied by applying spatial correlation analysis, such as global spatial autocorrelation analysis and hot spot analysis. Based on the study, the system of driving force indicators was constructed from the field of nature, society and economics according to “Regulations for valuation on agricultural land”, which includes the classification of agricultural land, gross domestic product (GDP), proportion of the first industry, public revenue, public expenditure, urbanization rate, per capita GDP, and rural per capita net income. Meanwhile, correlation analysis and spatial econometric models were used to reveal the spatial pattern of cultivated land transfer rent and its forming mechanism. The results showed that, in general, the cultivated land transfer rent in Henan Province presented patterns of apparent invert U in the east-west direction and gradually emerged a descending trend from north to south. In the global spatial autocorrelation analysis, Moran's index reached 0.63, which indicated that the cultivated land transfer rent showed a significant clustering phenomenon in space, and specifically, the space pattern showed a high-high and low-low clustering. Furthermore, by applying hot spot analysis, it was found that the hot spots and sub hot spots were intensively distributed in the piedmont plain of Taihang Mountains and eastern plain of Henan Province, the sub cold spots in the periphery of the sub hot spots, such as cities of Pingdingshan, Eastern Nanyang and Zhumadian, and the cold spots mostly in Funiu Mountainous region and Dabie Mountainous region. The results of correlation analysis indicated that driving factors, including the classification of agricultural land, GDP, public revenue, urbanization rate, and rural per capita net income, were significantly correlated with the cultivated land transfer rent, in which the classification of agricultural land showed a significant negative correlation with the cultivated land transfer rent, while GDP, public revenue, urbanization rate, and rural per capita net income were positively associated with it. Moreover, among above factors, the significant spatial coupling with the cultivated land transfer rent still existed in the classification of agricultural land, GDP, public revenue and rural per capita net income. Spatial Error Model (SEM) was proved to be more effective and robust when compared with Spatial Lag Model (SLM) in the process of spatial econometric analysis by comparing precision parameters, such as R2, LogL, AIC, and SC. In the Spatial Error Model, the classification of agricultural land, GDP and rural per capita net income could satisfy the regression accuracy of 0.01, and public revenue also reached up to 0.05 significant level. The result of collinearity diagnostics showed that multiple collinearity existed among GDP, public revenue and rural per capita net income, and explained the puzzle of the coefficient of GDP being negative in spatial econometric model. In the ensuing analysis, stepwise regression model could achieve optimal result when the classification of agricultural land and rural per capita net income were just introduced into the model. Based on the analysis above, it was reckoned that the cultivated land transfer rent was not randomly distributed but showed a strong spatial autocorrelation in space; the cultivated land transfer rent was markedly influenced by the natural and social economic factors; Compared with the GDP and public revenue, the effects of the classification of agricultural land and rural per capita net income on the cultivated land transfer rent were more significant.