OBJECTIVE: To identify sensitive periods of postnatal growth for preterm infants relative to neurodevelopment at 18 months' corrected age. PATIENTS AND METHODS: We studied 613 infants born at <33 weeks'gestation who participated in the DHA for Improvement of Neurodevelopmental Outcome trial. We calculated linear slopes of growth in weight, length, BMI, and head circumference from 1 week of age to term (40 weeks' postmenstrual age), term to 4 months, and 4 to 12 months, and we estimated their associations with Bayley Scales of Infant Development, 2nd Edition, Mental (MDI) and Psychomotor (PDI) Development Indexes in linear regression. RESULTS: The median gestational age was 30 (range: 2-33) weeks. Mean ± SD MDI was 94 ± 16, and PDI was 93 ± 16. From 1 week to term, greater weight gain (2.4 MDI points per z score [95% confidence interval (CI): 0.8 -3.9]; 2.7 PDI points [95% CI: 1.2-.2]), BMI gain (1.7 MDI points [95% CI: 0.4 -3.1]; 2.5 PDI points [95% CI: 1.2-3.9]), and head growth (1.4 MDI points [95% CI: -0.0 -2.8]; 2.5 PDI points [95% CI: 1.2-3.9]) were associated with higher scores. From term to 4 months, greater weight gain (1.7 points [95% CI: 0.2-3.1]) and linear growth (2.0 points [95% CI: 0.7-3.2]), but not BMI gain, were associated with higher PDI. From 4 to 12 months, none of the growth measures was associated with MDI or PDI score. CONCLUSIONS: In preterm infants, greater weight and BMI gain to term were associated with better neurodevelopmental outcomes. After term, greater weight gain was also associated with better outcomes, but increasing weight out of proportion to length did not confer additional benefit.