There are tensions within formal education between imparting knowledge and the development of skills for handling that knowledge. In the primary school sector, the latter can also be squeezed out of the curriculum by a focus on basic skills such as literacy and numeracy. What happens when an explicit attempt is made to develop young children's reasoning—both in terms of their apparent cognitive abilities and their basic skills? This paper reports an independent evaluation of an in‐class intervention called ‘Philosophy for Children’ (P4C), after just over one year of schooling. The intervention aims to help children become more willing and able to question, reason, construct arguments and collaborate with others. A group of 48 volunteer schools were randomised to receive P4C (22 schools) or act as a control for one year (26). This paper reports the CAT results for all pupils in years 4 and 5 initially, and the Key Stage 2 attainment in English and Maths for those starting in year 5. There was no school dropout. Individual attrition from a total of 3,159 pupils was around 11 percent—roughly equal between groups. There were small positive ‘effect’ sizes in favour of the P4C group in progress in reading (+0.12) and maths (+0.10), and even smaller perhaps negligible improvements in CAT scores (+0.07) and writing (+0.03). The results for the most disadvantaged (free school eligible) pupils were larger for attainment (+0.29 in reading, +0.17 writing and +0.20 maths), but not for CATs (–0.02). Observations and interviews suggest that the intervention was generally enjoyable and thought to be beneficial for pupil confidence. Our conclusion is that, for those wishing to improve attainment outcomes in the short term, an emphasis on developing reasoning is promising, especially for the poorest students, but perhaps not the most effective way forward. However, for those who value reasoning for its own sake, this evaluation demonstrates that using curriculum time in this way does not damage attainment (and may well enhance it and reduce the poverty gradient in attainment), and so suggests that something like P4C is an appropriate educational approach.