This paper surveys the growing literature on developing countries’ infrastructure performance measured through frontier analysis techniques. The focus is generally on assessments of operators’ efficiency and productivity growth as part of benchmarking exercises. Two broad types of benchmarks are identified in the literature. The first is based on cross-country or cross regional performance comparisons, the second compares performance before and after reforms, including privatization. These benchmarks are often used to assess the impact of potential or actual policy reforms in specific countries. In a growing number of instances, they are also being conducted in the context of tariff revisions to assess the efficiency gains that can be shared with users as part of these revisions. The main policy conclusions are that there is a difference in the performance effect of key reforms, in particular privatization, between the transport sector and the utilities sector but that generally incentive based regulation delivers efficiency in both sectors The main technical conclusion is that even if in developing economies data problems are constraining, these constraints are not as binding as often argued.