Agri-environmental payments are increasingly used to incentivize sustainable land use. The design of these payments needs to overcome two principal challenges. First, cooperation among land users is often required to achieve environmental outcomes. Second, conservation costs of adopting sustainable land use usually vary among farmers. Agglomeration payments have been proposed to address the first issue, while differentiated payments have been proposed to address the second. In practice, uniform payments are most commonly used and set equal to average opportunity costs. Past empirical studies focus either on the first or on the second issue. In this paper we combine the two. Using the issue of cultivated organic soils in Switzerland as specific example, we test the effectiveness, efficiency and equity of differently designed conservation payments in a computerized and interactive framed economic experiment that we conducted with university students acting as farmers. The experimental setup allows for side payments between players. We also collected data on individual social preferences and a variety of socio-demographic characteristics. We find that agglomeration payments differentiated by players' opportunity costs and uniform agglomeration payments equal to average players' opportunity costs are equally environmentally effective. This is because group members use side payments to redistribute payoffs among different cost types. In terms of cost effectiveness, differentiated payments perform best. Nevertheless, differentiated payments lead to highest income inequalities among players. Finally, we find a strong link between individual social preferences and behavior in the experiment.