Estimating the effect of inequality on crime is challenging due to reverse causality and omitted variable bias. This paper addresses these concerns by exploiting the fact that, as suggested by recent scholarly research, the legacy of slavery is largely manifested in persistent levels of economic inequality. Municipality-level economic inequality in Colombia is instrumented with a census-based measure of the proportion of slaves before the abolition of slavery in the nineteenth century. It is found that inequality increases both property crime and violent crime. The estimates are robust to including traditional determinants of crime (like population density, proportion of young males, average education level, quality of law enforcement institutions, and overall economic activity), as well as geographic characteristics that may be correlated with both the slave economy and with crime, and current ethnic differences. Policies aiming at reducing structural crime should focus on reducing economic inequality.