This study aimed to determine whether neonatal feeding performance can predict the neurodevelopmental outcome of infants at 18 months of age. We measured the expression and sucking pressures of 65 infants (32 males and 33 females, mean gestational age 37.8 weeks [SD 0.5]; range 35.1 to 42.7 weeks and mean birthweight 2722g [SD 92]) with feeding problems and assessed their neurodevelopmental outcome at 18 months of age. Their diagnoses varied from mild asphyxia and transient tachypnea to Chiari malformation. A neurological examination was performed at 40 to 42 weeks postmenstrual age by means of an Amiel-Tison examination. Feeding performance at 1 and 2 weeks after initiation of oral feeding was divided into four classes: class 1, no suction and weak expression; class 2, arrhythmic alternation of expression/suction and weak pressures; class 3, rhythmic alternation, but weak pressures; and class 4, rhythmic alternation with normal pressures. Neurodevelopmental outcome was evaluated with the Bayley Scales of Infant Development–II and was divided into four categories: severe disability, moderate delay, minor delay, and normal. We examined the brain ultrasound on the day of feeding assessment, and compared the prognostic value of ultrasound and feeding performance. There was a significant correlation between feeding assessment and neurodevelopmental outcome at 18 months (p<0.001). Improvements of feeding pattern at the second evaluation resulted in better neurodevelopmental outcome. The sensitivity and specificity of feeding assessment were higher than those of ultrasound assessment. Neonatal feeding performance is, therefore, of prognostic value in detecting future developmental problems.