A large literature concurs that social determinants of health (SDH) are demonstrable, important, and insufficiently attended to in policy and practice. A resulting priority for research should be to determine how the social determinants of health can best be addressed. In this paper we support the more effective transfer of social determinants research into policy by: (1) describing a qualitative analysis of thirty-two cancer control policy documents from six English-speaking OECD countries and two transnational organizations, demonstrating great variability in the treatment of social determinants in these policies; (2) critiquing these various policy practices in relation to their likely impact on social determinants of health; and (3) advancing a tool that policy writers can use to assess the way in which social determinants of health have been addressed in their work. In the sample of policy documents, the distinction between structural and intermediate determinants, population-based and targeted interventions, and their respective relationships to equity were not always clear. The authors identified four approaches to social determinants (acknowledging SDH, auditing SDH, stating aims regarding SDH and setting out actions on SDH), and five ways of writing about the relationship between social determinants and cancer risk. These five discourses implied, respectively: that group membership was intrinsically risky; that not enough was known about SDH; that risk arose from choices made by individuals; that groups were constrained by circumstance; or that structural change was necessary. Socio-cultural factors were generally presented negatively, though New Zealand policies modeled a possible alternative. Based on their empirical work, the authors propose a matrix and a set of questions to guide the development and assessment of health policy.