This paper addresses the nexus between evaluation of energy-efficiency programs and incentive payments based on performance for program administrators in California. The paper describes the problems that arise when evaluators are asked to measure program performance by answering the counterfactual question what would have happened in the absence of the program? Then the paper examines some ways of addressing these problems. Key conclusions are (1) program evaluation cannot precisely and accurately determine the counterfactual, there will always be substantial uncertainty, (2) given the current state of knowledge, the decision to tie all incentives to program outcomes is misguided, and (3) incentive programs should be regularly reviewed and revised so that they can be adapted to new conditions. (C) 2010 Elsevier Ltd. All rights reserved.