Urban agriculture, a current trend in many US cities, is purported to bring enhanced food security, reduction of food waste, community building, open green space in cities and higher property values. However, the literature lacks an understanding of whether urban farming has extended beyond a compelling concept into the practice of farming in the city and peri-urban areas. The exact definition of an urban farm is challenging, since many urban farms have a primary mission of supporting social goals rather than providing food. Use of the USDA definition of farm omits many self-identified urban farms, but the most consistent measure of agriculture is the Census of Agriculture. Using census data, this paper finds that urban farms are smaller than the typical farm, and while the amount of urban and peri-urban farmland declined between 2002 and 2007, the total number of farms increased. Growth in farmland is positively related to land values, suggesting that increases in urban farmland are more likely to take place in population dense, land scarce areas. Spatial analysis of urban and peri-urban farms in the Northeast finds fewer clusters of farms in areas with high land costs. In the most populous Northeastern cities, the farms are more likely to be located in the peri-urban area than in the urban core. Urban farms in the Northeast were more likely to produce vegetables, eggs and goats. Significant levels of vegetable farm clusters were detected surrounding Providence, Boston and Hartford Metropolitan Statistical Areas, which are regions that had no significant level of clustering of total farms. Future analysis, incorporating data from the 2012 census, should provide insight into whether local policy changes have resulted in growth in urban farms and farmland.