Objective: To examine performance in preschool and academic skills in very preterm (gestational age ≤30 weeks) and term-born comparison children aged 4 to 12 years. Study design: Very preterm children (n = 200; mean age, 8.2 ± 2.5 years) born between 1996 and 2004 were compared with 230 term-born children (mean age, 8.3 ± 2.3). The Dutch National Pupil Monitoring System was used to measure preschool numerical reasoning and early linguistics, and primary school simple and complex word reading, reading comprehension, spelling, and mathematics/arithmetic. With univariate analyses of variance, we assessed the effects of preterm birth on performance across grades and on grade retention. Results: In preschool, very preterm children performed comparably with term-born children in early linguistics, but perform more poorly (0.7 standard deviation [SD]) in numerical reasoning skills. In primary school, very preterm children scored 0.3 SD lower in complex word reading and 0.6 SD lower in mathematics/arithmetic, but performed comparably with peers in reading comprehension and spelling. They had a higher grade repeat rate (25.5%), although grade repeat did not improve their academic skills. Conclusions: Very preterm children do well in early linguistics, reading comprehension, and spelling, but have clinically significant deficits in numerical reasoning skills and mathematics/arithmetic, which persist with time. © 2011 Mosby Inc. All rights reserved.