The relationship between deprivation and educational outcomes has been the subject of a long-running and deep debate in the economic literature. Recent discussions have focused on causality, with experimental and quasi-experimental approaches taken, yet, predominantly, the literature continues to proxy deprivation with measures of wealth. This paper explores a much wider measure and identifies a causal relationship between regional deprivation and school performance in Northern Ireland. Combining panel data on Key Stage II results from each of Northern Ireland’s primary schools with the 2005 Northern Ireland Multiple Deprivation Measure, we show the net negative impact of this wider measure, whilst an extension explores the impacts of each single domain. Using an error-component two-stage least squares model, we account for school and neighbourhood selection and the potential endogeneity of our deprivation measure, showing spatial variation in historical violence, which occurred during “The Troubles”, to be a valid instrument for deprivation. Our results confirm the negative impact of deprivation frequently found in the literature but also that, when the impacts of other deprivation domains are accounted for, education and crime deprivation, and not financial deprivation, play a significant role in determining outcomes. This confirms the limitations of using wealth as a proxy for neighbourhood deprivation, whilst suggesting that policies focusing only on income redistribution will be unsuccessful in improving education outcomes of those exposed to deprivation.