Reducing access to lethal means can prevent suicides. However, substitution of a suicide method remains a concern. Until 1986, the Ellington Bridge was the site of one-half of all Washington, DC bridge suicides. An antisuicide fence was installed in 1986, creating a naturalistic case–control design for testing the substitution hypothesis with the adjacent and equally as lethal jump site, the Taft Bridge. We found that suicide deaths from the Ellington Bridge were reduced by 90% (p=0.001) following barrier construction, without changes in rates of jumps from either the Taft Bridge or any other bridge in the city. Suicides by all methods decreased significantly across the study period. While the decline in suicides from the Ellington Bridge may reflect a broader decline in suicide, the decline in bridge suicide without persistent shifts in deaths to other bridges provides evidence that restricting access to one highly lethal method is effective.