In this article, we study the performance of international water management at Lake Titicaca in order to assess the empirical applicability of a new concept for measuring policy performance. This performance measurement concept (PER) is a function of three variables: optimum performance (OP), actual performance (AP), and counterfactual performance (CP). In examining the joint management of Lake Titicaca by Bolivia and Peru, we identify practical difficulties in applying PER and suggest extensions of the concept to assess uncertainty. We find that this measurement concept has several advantages: it takes into account counterfactual performance; the analysis can be performed in a structured, standardized, and transparent manner; and the formal framework allows for an assessment of uncertainty. Problems in applying the concept include: choice of attributes, identification of optimum performance, time-scale, and subjectivity of information. We conclude that, on balance, this approach is superior to other policy performance assessment measures developed to date.