A group of 347 children who had been classed as 'at risk' in the neonatal period were examined by means of academic attainment tests and teacher ratings at the end of their first year in primary school. Matched classroom controls were also examined to estimate the prevalence of reading, spelling and mathematics difficulties. Preterm birth and low birthweight appeared to be significant antecedents of poorer outcome in all three subjects. The attributable risk of low birthweight to academic handicap was estimated to be 29 per cent. There was no additional effect of time to spontaneous respiration (short and long) on academic outcome for the preterm low-birthweight group. In contrast, poor outcome for the low-birthweight children who had not been preterm was associated with longer time to spontaneous respiration. Graduates of neonatal intensive care with normal birthweights performed comparably with their low-risk controls. These results are discussed in relation to other longitudinal studies of the outcome for 'at-risk' infants.