The purpose of this paper is to analyze the differentials and detect factors affecting infant mortality on the basis of data obtained from the fertility surveys from those countries participating in the World Fertility Survey. In particular, this includes the surveys carried out in Colombia, Peru, Costa Rica, Panama, and the Dominican Republic. 3 types of explanatory variables may be considered from the information available: 1) context variables related to the mother's environment; 2) socioeconomic variables based on the educational and economic characteristics of the mother and her last husband; and 3) biological factors (from each woman's pregnancy history) such as mother's age at birth of the child, order of birth, interbirth interval, etc. The countries, whether high or low mortality, present great differences in child mortality in most of the variables considered. In Panama and Costa Rica there are population sectors with infant mortality rates of around 100/1000 live births, whereas in Peru these are over 150/1000 (children from mothers without education, low agricultural strata, etc.). Besides presenting the differentials, a methodological test is made through the application to Costa Rica and Peru of the Proportional Hazards Model which permits analysis of the effects of variables when acting simultaneously upon mortality in early childhood. The variables which show the highest disparity in mortality level are: natural region among the context variables, education of mother among the socioeconomic variables, and interbirth interval and maternal age at birth of their children among the biological ones.