The Energy Star labeling program contributes to reducing economic and environmental impacts. Although product-specific criteria for the label are developed by taking into account potential energy savings, more robust criteria need to be developed to minimize energy consumption (and consequential global warming potential (GWP)). The objectives of this study are: (i) to measure what we refer to as the technical and cost efficiencies of products with respect to the relative Energy Star criteria achievement rate and GWP, respectively; and (ii) to analyze the relationship between these efficiencies and product design features, using stochastic frontier analysis (SFA) and data envelopment analysis (DEA). As an example, these analyses are applied to data for desktop computers. The results indicate that desktop computer manufacturers currently achieve energy savings that exceed those required by the Energy Star label. The principal product design features affecting the efficiencies are primarily the efficiency and power capacity of the power supply and the number of cores in the central processing unit. This study can help the program develop more robust criteria that ultimately reduce GWP and can provide guidance to manufacturers to strategically improve energy efficiency. (C) 2011 Elsevier Ltd. All rights reserved.