Due to high losses of agricultural land in urbanizing areas over the past several years, state and local governments have adopted zoning regulations, right-to-farm ordinances, preferential property tax programs, and other means to protect agricultural land resources. Critics of the farmland preservation efforts note that these programs may only delay the ultimate conversion of farmland to urban uses and may simply shift urban development to neighboring areas (positive spillover effects). An alternative means of permanently protecting farmland is provided by purchase of development rights (PDR) programs, which enroll acreage from landowners who voluntarily sell the development rights to their farmland. Although ownership and all other property rights are retained by the owner, the option to develop the land for urban uses is removed from the current and all subsequent landowners. The purpose of this paper is to develop an empirical model of the impact of PDR programs on urban land development rates. To estimate the empirical model, we form a dependent variable based on county level rates of urban development from the USDA National Resources Inventory (NRI) data base for nine states in the Northeast US that have adopted farmland protection measures during the sample period (1982-1997). The set of explanatory variables include dummy variables for the existence of an active PDR program in the county and active PDR programs in adjacent counties plus measures of demographics, farm returns, population growth, income, and site-specific factors gathered from several sources (e.g., Census of Agriculture, Census of Population, and USDA). The model is estimated using the ordinary least squares technique. Two key policy-relevant hypotheses: (1) PDR programs did not reduce the rate of urban development and (2) PDR programs did not induce positive spillover effects on neighboring counties. The test results suggest that PDRs may have had important more significant impact in the earlier years and did not cause positive spillover effects.