Payments for ecosystem services (PES) are emerging worldwide as important mechanisms to align investments in human and natural well-being. PES projects are often defined as voluntary transactions where well-defined environmental/ecosystem services (or land uses likely to secure those services) are bought by a minimum of one service buyer, from a minimum of one service provider, if and only if the service provider continuously secures service provision (conditionality). Further criteria of PES include limiting additional objectives and ensuring that payments reward behaviour that would otherwise not occur (additionality). Together these best practices for PES are increasingly accepted as the most efficient means to achieve desired outcomes and are guiding funding for PES projects. We used a series of water funds (watershed-oriented PES projects based on a trust fund model) to examine how theoretical best practices could inform and improve practice and also how theory could learn from practical efforts. We conclude that thoughtful consideration is required when evaluating the promise of a PES approach against a theoretical ideal. We found that requiring conditionality may limit the use of creative finance mechanisms such as trust funds that can provide long-term benefits for conservation and human well-being, and that requiring additionality can exclude benefits from social diffusion and result in the inefficient targeting of PES funds. Finally, public–private partnerships in water funds lead to multiple additional/side objectives but partnerships are likely to lower transaction costs and provide transparent, long-term landscape-scale watershed management.