Large-scale environmental programs generally commit substantial societal resources, making the evaluation of their actual effects on the relevant outcomes imperative. As the example of the subsidization of energy-saving appliances illustrates, much of the applied environmental economics literature has yet to confront the problem of proper attribution of effects to underlying causes on a convincing methodological basis. This paper argues that recent results in the econometrics and statistics literature on program evaluation could be utilized to advance considerably in this context. In particular, the construction of a credible counterfactual situation is at the heart of the formal statistical evaluation problem. Even when controlled experiments are not a viable option, appropriate approaches might succeed where traditional empirical strategies fail to uncover the effects of environmental interventions.