This paper examines how the quality of early childhood education and care accessed by 3- and 4-year-olds in England varies by children's background. Focusing on the free entitlement to early education, the analysis combines information from three administrative datasets for 2010-2011, the Early Years Census, the Schools Census and the Ofsted inspections dataset, to obtain two main indicators of quality: staff qualification levels and Ofsted ratings. These data are combined with child-level indicators of area deprivation (IDACI scores) as a proxy measure of children's background. The paper finds that children from more disadvantaged areas have access to better qualified staff, largely because they are more likely than children from richer areas to attend maintained nursery classes staffed by teachers, and less likely to attend services in the private, voluntary and independent (PVI) sectors. However, "within" both maintained and PVI sectors, services catering for more disadvantaged children receive poorer quality ratings from Ofsted, with a higher concentration of children from disadvantaged areas itself appearing to reduce the likelihood of top Ofsted grades. This may be in part because Ofsted ratings reflect levels of child development, and therefore reward settings where children enter at a more advanced starting point, but it may also be that it is genuinely harder to deliver an outstanding service to a more disadvantaged intake. The results point to the need for funding to support better qualified staff in PVI settings in disadvantaged areas.