One of the Millennium Development Goals is the reduction of infant and child mortality by two-thirds by 2015. In order to achieve this goal, efforts are concentrated at identifying cost-effective strategies as many international agencies have advocated for more resources to be directed to health sector. One way of doing this is to identify and rank-order the importance of the socioeconomic factors that affect infant mortality. This will help in prioritizing the factors that need to be manipulated for effective health interventions in the face of competing scarce resources. The purpose of this study is to examine the relative importance of major biosocial, demographic and economic factors associated with infant mortality in Kenya. The study is an analytical cross-sectional design through secondary data analysis of the 2003 Kenyan Demographic and Health Survey (KDHS) dataset for children. Series of logistic regression models were fitted to select the significant factors affecting infant mortality both in urban and rural. The magnitude of the significance for each selected variable was tested using the Wald’s test, and hence the factors were rankordered according to their overall P-value. While infant mortality rate in 2003 was 79.6 per 1000, its significant determinants are breastfeeding, ethnicity and sex of the child while birth order and intervals are significant variables in the rural areas. Focus of interventions in child health with a view to achieving the MDG will be on the social and economic empowerment of women via education and employment while breastfeeding promotion will be encouraged