Childhood vaccination rates have plateaued and, in some cases, declined—raising concerns about disease resurgence and school health. While family, cultural, and access factors are known determinants of vaccine uptake, less is understood about peer influences within schools. This paper examines peer effects in vaccination behavior using the staggered approval of COVID-19 vaccines for children ages 5–17 in 2021. Exploiting exogenous variation in vaccine eligibility determined by age cutoffs, we estimate the causal impact of peers’ eligibility and vaccination on individual take-up and absenteeism among 170,543 New York City middle school students across 467 public schools. Our identification strategy instruments schoolmate and grade-mate vaccination rates with corresponding eligibility rates, leveraging differences in age composition across and within schools. Results show strong peer effects: a one standard deviation increase in grade-mate vaccination raises individual take-up by 5.6 percentage points, compared to 1.9 points for schoolmates. Peer vaccination also improves attendance—raising attendance rates by 1.5 percentage points and reducing chronic absenteeism (≤90% attendance) by 5.8 percentage points—effects concentrated among Black and Hispanic students. COVID vaccination, while not mandatory, was highly salient and uncertain, offering a rare setting to observe endogenous peer influence in a public health context. The findings highlight the importance of social learning and norm formation in vaccine uptake and suggest that peer-targeted school interventions may improve both immunization and academic outcomes, offering broader lessons for other voluntary vaccines such as influenza.