This paper considers the effect of how children pass time before entrance to school on attainment in primary school. We find in National Child Developement Study data that Children perform marginally better at 7 and 11 if they spent time with their mother, or at a prre-school, rather than in informal care. This holds when one controls for parental education, social class and assessed parental interest in the child's education, as well as the quality of the peer group. In the British Cohort Study, however, time spent in nurseries effected no improvement in mathematics at 10 as compared with time in informal care, and pre-school children were performing much worse in reading. This worse performance was traceable to reduced vocabulary at 5. Pre-school children were more advanced in copying at 5 relative to children in informal care, but, while copying is a good predictor of scores in both mathematics and reading at 10, this advancement had been offset by then.