This study utilizes population-level birth data to describe those children who may be at greatest risk of maltreatment during the first five years of life. Based on a unique dataset constructed by linking California’s administrative child welfare data to statewide vital birth records, a cohort study design was employed to track reports of maltreatment involving children born in 2002. Twelve variables captured in the birth record were selected for analysis. Generalized Linear Models were used to estimate adjusted risk ratios (RR) for each independent variable. Predicted probabilities of CPS contact were computed based on the count of risk factors present at birth. Results suggest that many of the associations previously observed between birth variables and subsequent maltreatment have sustained value in foretelling which children will be reported to CPS beyond infancy. Of the 531,035 children born in California in 2002, 14% (74,182) were reported for possible maltreatment before the age of five. Eleven of the twelve birth variables examined presented as significant predictors of contact with child protective services.