How do local news outlets cover crime, and what are the effects of said coverage on housing markets? By linking almost the universe of individual homicides to individual stories in the news, this paper estimates the bias in coverage of crime news by local TV outlets in the United States. This paper finds evidence of substantial biases in the coverage of homicides along race (victim and suspect) and wealth: inter-racial crime is substantially over-covered; white-on-white crime is under-covered; and richer neighborhoods get substantially more coverage. I find that race bias depends on the media market demographics, with stations in markets with a higher Black share having less over-coverage of inter-racial homicides and higher coverage of nonwhite intra- race crimes. I also find that, controlling for market size, within-market variation of bias increases with the number of stations. I propose a simple model of horizontal differentiation that captures these results, suggesting that stations are catering to the largest demographic group. Lastly, I evaluate the impact of crime coverage on housing prices using an instrumental variable strategy based on sporting events, which crowd out crime news. The results show that bias in crime coverage puts downward pressure on housing prices in minority neighborhoods.