This paper tests the common belief that subsidised housing contributes to higher crime rates. To do this, panel data on over 200 US cities are used and fixed effects models are estimated to control for unobserved differences between cities that may affect both voucher use and crime. Additionally, models are estimated that focus on the suburbs, to see if the steady increase in vouchers there has had any effect on crime. In cities, it is found that vouchers have a weak, negative relationship with violent crime rates, although these estimates are not particularly robust. In suburban areas, there is no observed relationship between vouchers and crime, suggesting that controversies in those communities blaming voucher households for elevated crime rates are misguided.