The value of ability grouping is often debated despite being adopted in primary and secondary schools across the UK for the past 80 years. Setting is one form of ability grouping which is widely adopted in English schools; it involves dividing pupils from the same cohort into classes according to ability in a specific subject. While the existing evidence identifies a negative effect on cognitive outcomes, especially for low ability pupils, little research has been undertaken to understand the impact of setting on non-cognitive outcomes. This paper provides the first evidence of the effect of setting on non-cognitive outcomes when utilising a nationally representative sample of primary-aged pupils and adopting fixed effects and instrumental variables methodologies. For boys, setting in maths negatively impacts non-cognitive outcomes, driven by a worsening of internalising behaviours. No evidence of a significant impact of lowest set placement on non-cognitive outcomes is identified.