This study explores crime underreporting in the United States in 2013. In analyzing this phenomenon, it seeks to isolate the factors that cause underreporting, as well as certain reasons why underreporting occurs. I use data from the 2013 National Crime Victimization Survey to estimate three separate models. The first model uses a probit regression to isolate several factors that cause crime underreporting. The second model uses a probit regression with inverse probability weighting to correct for selection bias for crime victims. The third model uses a multinomial logit to analyze the most prevalent reasons for underreporting. The results of the first regression suggest that crime type, knowing the perpetrator, age, gender, marital status, and income are the most prevalent factors that affect underreporting. The results of the second regression illustrate the need for selection correction, and point to crime type, gender, marital status, income, and race as the most important factors affecting crime underreporting. The results of the final regression suggest several relationships between underreporting and reasoning, including a link between assault, knowing the perpetrator, and dealing with the crime another way, a link between age and not reporting a crime because it is too insignificant to report, and a link between being Hispanic and not reporting a crime because the police would not help the situation. These findings are presented as a guide to policymakers on the areas they need to address to curb the problem of crime underreporting.