It has been proposed that cross-country comparisons of the technical efficiency of health production, estimated using data envelopment analysis (DEA), have useful applications for policy makers. In theory such an analysis utilizes measures of the socioeconomic determinants of health relevant to all social policy, not just health policy. Using OECD and WHO data, this article critically analyses a number of outstanding theoretical questions regarding the use of DEA in this setting. It concludes that until such questions are addressed, the resultant implications for policy will be based on misleading information.