The discrepancy in official statistics regarding feminicide in Peru –a variable that has included only feminicide and not attempted feminicide, in spite of the thin line that divides them in terms of the aggressor’s intentionality– has led to an incorrect and partial reading of this phenomenon. To overcome this problem, we built the variable aggregated feminicide (the sum of feminicide and attempted feminicide) and found it has been growing constantly and at increasing rates over time. On this basis, the objective of this paper was to study the macrosocial determinants of aggregated feminicide. Using a panel data econometric model (2009-2013), we obtained evidence of a stronger association between aggregated feminicide and protective factors (education, economic autonomy, and accessibility to health establishments) than between aggregated feminicide and risk factors (physical violence). The policy implications point to the need for developing macrosocial indicators to improve the focalization and effectiveness of public policy regarding feminicide and violence against women.