In response to the Every Student Succeeds Act (2015), nearly three-fourths of states in the U.S. have adopted chronic absenteeism—defined as missing 10% of the school year—as a measure of school quality and student success (Jordon, Fothergill, & Rosende, 2018). Due to its widespread adoption and the strong predictive relationship between early absences and negative educational outcomes, chronic absenteeism is increasingly being utilized by schools as an early warning indicator of later problems, such as low academic achievement. As such, chronic absenteeism theoretically allows schools to identify academically at-risk students in the early primary grades using readily available attendance data and provide them with additional resources to prevent later difficulties (Balfanz, Herzog, & Mac Iver, 2007). Given its pervasive use as both an accountability metric and an early warning indicator, the need to ensure the scientific integrity of chronic absenteeism is vital. Major theoretical assumptions underlying this indicator, however, have never been empirically validated. The current study represents the first effort to scientifically test the most basic assumption upon which chronic absenteeism is based—that all absences from school (i.e., both excused and unexcused absences) are equally detrimental to student outcomes and should be utilized to identify at-risk students. The purpose of this study was thus to test whether excused and unexcused absences have comparable diagnostic accuracy in the early identification of academically at-risk students. Using the state-of-the-art receiver operating characteristic (ROC) methodology, this study presented evidence that only unexcused absences provided diagnostic accuracy for academic risk status in math and English achievement for an entire cohort of young students in Philadelphia. This diagnostic accuracy was evident in kindergarten and increased across the early elementary years. Excused absences, on the other hand, provided no diagnostic utility in differentiating between students at risk for academic problems and students on track for success within and across the early elementary grades. The findings presented here indicate that chronic absenteeism could be a more effective early warning indicator for students in large urban school districts by taking absence types into account. These results have further implications for researchers and policymakers, surfacing the need to prioritize additional empirical studies testing the underlying assumptions of chronic absenteeism.