By examining the actual perfommance of conservation or demand-side management (DSM) programs for ten utilities, Joskow and Marron (1992) have made an important contribution to policy discussions about the wisdom of relying on utilities to improve customer energy efficiency. We use Joskow and Marron＇s method to analyze twenty utility commercial lighting programs and, like Joskow and Marron, find wide variations in industry reporting practices and savings evaluation methods. We extend the method by systematically accounting for several of the most important sources of variation and comment on how they influence total program costs. Our accounting also allows us to relate remaining program cost variations to the program sizes and the electric supply costs avoided by the programs. We draw qualified, yet affirmative, conclusions regarding the cost effectiveness of the programs.