Infant mortality rates (IMR) have traditionally been considered useful as health status indicators, and changes in these rates are thought to reflect changes in both medical care services and socio-economic circumstances. In order to explore this relationship of IMR with medical care and socio-economic factors in a developing country, Chilean health zone data for the decade 1960–1970 were used to construct 25 variables which were then classified into groups representing antenatal-obstetric services, acute and preventive medical services and socio-economic variables. In an analysis which involved developing a series of linear multiple regression equations for each year of the decade 1960–1970 with IMR as the dependent variable, the percentage of births with professional attention proved to be the strongest variable.