Policymakers and educators continue to closely track post-COVID student recovery, and with COVID-19 relief funds set to expire, decision makers need clear answers and actionable insights. Yet, accurate depictions of student performance have been challenging to gather, given the multifaceted and varied ways academic performance is assessed and these data reported. Obtaining more accurate depictions of student performance requires both longitudinal data—to examine the consequences of COVID-19 (including COVID learning loss), school closures, and the related recovery efforts—and disaggregated data—to examine how results vary by student, school, and community characteristics. Leveraging three years of longitudinal data in comparison to historic growth patterns reveals disparate trends by student age during the pandemic, prior achievement, and school and community characteristics. We find younger students and students who were already needing support academically are the furthest away from reaching pre-pandemic growth trends. In parallel, students from schools serving lower-income or minoritized communities continue to perform below pre-pandemic trends. These data suggest some populations or communities have benefited more, while those most in need of support continue to need it. Results bring into question whether intervention approaches have been as targeted as needed or whether there is a misalignment between the interventions tried and the populations served.