In stated preference assessments of farmland preservation programs, respondents are often told that preservation will occur within a given scale-e.g., community, state, county-but do not know the specific location of parcels in question. Hence, welfare estimates may be available for different scales, providing numerous avenues for benefit transfer. This paper provides a systematic assessment of transfer error, contrasting methods for the transfer of farmland preservation values across states and jurisdictional scales. The data are drawn from choice experiments conducted simultaneously in two different states and at two different scales. Results suggest that transfers across state outperform transfers across scale, and that simpler methods often outperform more complex approaches that calibrate for differences across scale.