We investigate the degree to which youth criminal offenses are influenced by neighbors, identifying causal effects with a natural experimental allocation of social housing in Copenhagen.  We find that youth exposed to a one percentage point higher concentration of neighbors with drug criminal records are 6% more likely to be charged for criminal offenses, and that this impact manifests itself after six months of exposure. This neighborhood effect is higher on previous offenders, and does not lead to criminal partnerships. Our exploration of alternative mechanisms suggests youth interaction in proximate residential context with adults with drug crime experience as the most plausible source of neighborhood effects. Joint with George Charles Galster (Wayne State University).