We develop a theory of change for payments for environmental services (PES) to review their imminent strengths and weaknesses in light of a growing body of impact evaluation studies. We show that PES are probably at least as environmentally additional as other conservation tools, based on the limited evidence. The original vision of PES as being direct, flexible, and potentially effective remains valid, but PES design and implementation have to be upgraded in their economic functioning to better realize this potential. Adverse self-selection, inadequate administrative targeting, and ill-enforced conditionality constitute three key obstacles that may considerably hamper PES success. Policies such as spatial targeting to service density, threat and cost levels, and payment differentiation can alleviate the design challenges. PES site selection needs to further move into high-threat areas. Making adequate PES design choices also requires the political will to boost environmental effects.