Payments for Ecosystem Services (PES) are a widely used approach for forest conservation through which people are paid to avoid deforesting land they enroll in the program. We present findings from a randomized trial in Mexico that tested whether a PES contract that requires enrollees to enroll all of their forest is more effective than the traditional PES contract that allows them to exercise choice. The modification's aim is to prevent landowners from enrolling only parcels they planned to conserve anyway while leaving aside other parcels to deforest. We find that the full-enrollment treatment reduces deforestation by 41% compared to the traditional contract. This extra conservation occurs despite the full-enrollment provision reducing the compliance rate due to its more stringent requirements. The full-enrollment treatment more than quadrupled cost-effectiveness, highlighting the potential to substantially improve the efficacy of conservation payments through simple contract modifications.