Many schools across the United States use the Early Warning System (EWS) to identify chronically absent students on time and intervene. The prediction power of the EWS is well examined but more evidence is needed about their effectiveness at reducing student absence. This study examines the effect of EWS on student absence in a large urban school district using a multiple-cutoff regression discontinuity design. Findings indicate that EWS reduces chronic absence among socioeconomically advantaged students. However, it has no significant effect on chronic absence among socioeconomically disadvantaged students. Furthermore, EWS has no significant effect on moderate absence. These results suggest that schools should consider social and institutional barriers to improving attendance, especially among socioeconomically disadvantaged students.