In Peru, the discrepancy in the official figures on femicide — a variable which has included only femicides and not attempts despite the narrow line separating them in terms of the intention of the victim — has led to an incorrect and partial reading of the phenomenon. To overcome this problem, we construct the aggregate feminicide variable (summing of feminicides and attempts), and we identify its steady and increasing increase over time. On that basis, the aim of this article was to study the macrosocial determinants of aggregate femicide. Based on a panel econometric model of data (2009-2013), gain a stronger association between aggregated femicide and protective factors (education, economic autonomy and accessibility to health services) than between aggregated femicide and risk factors (physical violence). Policy implications point to the need to develop macro-social indicators to improve the targeting and effectiveness of public policies on femicide and violence against women.