Demographers have long studied the unacceptably high rates of infant mortality in the U.S., relative to other advanced countries. These higher rates are largely attributable to persistent racial gaps in infant health outcomes and are likely a reflection of social inequalities, which manifest as poorer infant health outcomes in certain groups. We extend on previous research in this area by utilizing the restricted 1990 and 2000 cohort linked birth-death files to examine the risk of infant death due to several main causes including: maternal complications, low birth weight, and other causes, and how it has changed over time. We estimate multinomial logistic regression models stratified by year to determine the likelihood of cause-specific infant death versus remaining alive. Our preliminary findings confirm an increasing black-white infant mortality gap across multiple causes of death. Further analysis will incorporate the most recent data available (2008).