As 70% of the killings of women takes place in the context of relational dynamics and in 80% of the cases the perpetrator is a man, we can presume that femicide constitutes much of the homicide mortality among women. Epidemiological surveillance of the killings of women can, therefore, provide indicators on the trends and geographical distribution of femicide and, indirectly, of the more general phenomenon of harassment and violence against women. The analysis of 40 years of mortality shows only a slight decrease of the murders of women nationwide. This suggests that the factors that underline this phenomenon are deeply rooted in the relationship between men and women. The decrease has taken place mainly in the South and Islands and the percentage SMRs point to a reversal of the relationship between geographic areas: thus, at the end of the observation period the North-West assumes a greater weight than the South and Islands. So we cannot exclude that part of the decrease in murders of women can be attributed to the overall decrease in homicides related to criminal activity, most pronounced in the South and Islands.