The effect of macro processes on infant mortality rates is explored in this analysis of Chilean provincial statistics over five decades. Urbanization and pluralism, as measured by the percent of the population in urban centers and the percent voting, predict lower infant mortality strongly and consistently. The theoretical rationale for linking macrostructural variables to biological outcomes is then examined in more detail. It is argued that the structural model employed here is an improvement on the conventional modernization/biomedical explanation because of its greater consistency and specificity, and because inclusion of measures of health technology in the tests did not eliminate the effect of the structural variables.