Previous research has found that the manner in which forests are managed reveals forest owners' goals and management preferences for their properties. In this study, we attempt to predict forest owners' participation in payments for ecosystem services (PES) programs using observable forest management in southeast Georgia, U.S. For this task, we rely on forest property characteristics identified from publicly available aerial images and databases. We examine whether and how forest property characteristics are related to family forest owners' preferences towards PES programs. The results from a random parameter logit model indicate that forest owners in southeast Georgia with planted pine forests, bottomland hardwood forests, or residential structures are more likely to participate in hypothetical PES programs. The results also indicate that forest owners' willingness to accept compensation to participate in PES programs is negatively related to the proportion of planted pine and bottomland hardwood forest areas, and to the existence of a residential structure on one's property. Findings of our study indicate the potential of revealed forest property characteristics in identifying and targeting PES programs.