OBJECTIVE: Among extremely preterm infants, we evaluated whether bevacizumab therapy compared with surgery for retinopathy of prematurity (ROP) is associated with adverse outcomes in early childhood. METHODS: This study was a retrospective analysis of prospectively collected data on preterm (22-26 1 6/7 weeks' gestational age) infants admitted to the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network centers who received bevacizumab or surgery exclusively for ROP. The primary outcome was death or severe neurodevelopmental impairment (NDI) at 18 to 26 months' corrected age (Bayley Scales of Infant and Toddler Development, Third Edition cognitive or motor composite score ,70, Gross Motor Functional Classification Scale level $2, bilateral blindness or hearing impairment). RESULTS: The cohort (N = 405; 214 [53%] boys; median [interquartile range] gestational age: 24.6 [23.9-25.3] weeks) included 181 (45%) infants who received bevacizumab and 224 (55%) who underwent ROP surgery. Infants treated with bevacizumab had a lower median (interquartile range) birth weight (640 [541-709] vs 660 [572.5-750] g; P = .02) and longer durations of conventional ventilation (35 [21-58] vs 33 [18-49] days; P = .04) and supplemental oxygen (112 [94-120] vs 105 [84.5-120] days; P = .01). Death or severe NDI (adjusted odds ratio [aOR] 1.42; 95% confidence interval [CI] 0.94 to 2.14) and severe NDI (aOR 1.14; 95% CI 0.76 to 1.70) did not differ between groups. Odds of death (aOR 2.54 [95% CI 1.42 to 4.55]; P = .002), a cognitive score ,85