[Objective] The carbon effect of urban agricultural land was estimated in Nanjing City to provide development experience for other cities with urban agriculture.[Methods] The carbon emissions from agricultural land in Nanjing City were calculated from 1996 to 2014 based on six main carbon sources of chemical fertilizers, pesticides, agricultural film, agricultural diesel, irrigation and tillage. At the same time, the variations of three main carbon sinks, including forest land, grassland, garden, were explored to illustrate the carbon effects of agricultural land use changes. In the end, the trend of carbon emissions from agricultural land in Nanjing City was predicted based on the gray GM(1,1) model.[Results] From 1996 to 2014, the carbon emissions from agricultural land use in Nanjing City showed a three-stage variation of "rising-fluctuating-steady decline". Chronically, the carbon sinks of both forest land and grass land decreased from 2005 to 2014, but their decreasing rate was different. From the view of regional comparison, due to the city carrying function was different, regional carbon sinks differed greatly. In terms of the carbon effect of the change of agricultural land use, the carbon sequestration in Nanjing City from 2000 to 2008 increased at first and then decreased; and carbon emission caused by construction occupancy experienced a Z-shaped variation. Based on the gray GM(1,1) model, the isobar metric forecasting method was used. It was estimated that the agricultural carbon use in the city will be 1.11×10⁵ t by 2020.[Conclusion] The trend of carbon use in Nanjing City was closely related to the economic development and the process of urban agriculture development. The reductions of forest land and grass land area reduced the original carbon sink effect. And as a result from economic development and urban construction needs, the continued growth in construction land have great impact, which will result in a large number of carbon emissions in Nanjing City.