This paper investigates the role of residential neighborhood context for juvenile crime. The paper estimates the neighborhood effect of potential offending neighbors on the criminal behavior of young neighbors at risk. To do so, the paper exploits the conditional quasi-randomization of first municipal social housing offer in Copenhagen. We find strong evidence that the share of neighbors with charges for drug-related crimes increase subsequent crime of juveniles, especially adolescents, in the 2 years following municipal housing. The seminar is a lunch seminar - please sign up by e-mail to mvandersen@econ.au.dk. Lunch is on the house.