The human losses after strong earthquakes that occurred in the world during the twentieth century have been analyzed, and a quantitative model for a preliminary assessment of casualties is proposed. It consists of a correlation between the number of casualties and the earthquake magnitude as a function of population density. We tackle the distribution of the total number of casualties within areas of different macroseismic intensity. Prognostic estimations of the expected number of killed or injured people caused by a supposed strong earthquake in Andalucia (Spain), using the model based on worldwide data, are suggested. Prognostic estimations based on specific data about the Kanto–Tokai (Japan) region are likewise given and compared with the number of casualties due to the 1995 Kobe (Japan) earthquake. In relation to the expected number of victims in areas affected by strong seismic impacts, we compute the casualty rate as the number of people killed divided into the inhabitants of a region and show its variation for different population density groups in the case of two extreme earthquake magnitudes.