The adoption of setting in the primary school (pupils ability grouped across classes for particular subjects) emerged during the 1990s as a means to raise standards. Recent research based on 8875 children in the Millennium Cohort Study showed that 25.8% of children in Year 2 were set for literacy and mathematics and a further 11.2% of children were set for mathematics or literacy alone. Logistic regression analysis showed that the best predictors of being in the top set for literacy or mathematics were whether the child was born in the Autumn or Winter and cognitive ability scores. Boys were significantly more likely than girls to be in the bottom literacy set. Family circumstances held less importance for setting placement compared with the child's own characteristics, although they were more important in relation to bottom set placement. Children in bottom sets were significantly more likely to be part of a long-term single parent household, have experienced poverty, and not to have a mother with qualifications at NVQ3 or higher levels. The findings are discussed in relation to earlier research and the implications for schools are set out.Keywords: ability grouping; longitudinal studies; school policy; primary education Introduction Grouping pupils by ability within and between schools is controversial and has been the subject of research since the early 1900s. In the primary school, grouping can take several forms: streaming (children are placed in classes on the basis of their perceived general ability); setting (children are grouped across classes for particular subjects based on their perceived ability in that subject); and within class (the teacher organises small groups based on ability within the class). These are not mutually exclusive and children may be streamed, placed in sets and working in within class groupings at any one time.The evidence from primary and secondary education suggests that, overall, structured ability grouping (streaming and setting), of itself, has no positive impact on average attainment, although, depending on the level of curriculum differentiation, can widen the gap between low and high attainers (e.g. Ireson et al. 2002;Kerckhoff 1986;Schofield 2010;Wiliam and Bartholomew 2004). It has personal consequences for pupils, particularly those in the middle and lower groups, as it *Corresponding author. Email: s.hallam@ioe.ac.ukResearch Papers in Education, 2014 Vol. 28, No. 4, 393-420, http://dx.doi.org/10.1080/02671522.2012 Ó 2012 The Author(s). Published by Routledge This is an Open Access article. Non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly attributed, cited, and is not altered, transformed, or built upon in any way, is permitted. The moral rights of the named author(s) have been asserted. can limit potential educational opportunities (e.g. Ball 1981;Boaler 1997;Lacey 1970), not only affecting expectations but in England setting very real limits on examination entry and possible attainment (Boaler, Wiliam, and B...