The highly contentious and partisan debate over voter identification reforms has not been well-informed by systematic evaluation studies of the impacts of such reforms on voter turnout.  In part this is because previous studies yield mixed findings, but also because the simplistic evaluation methods of earlier research. For example, previous studies have not accounted for differential identification requirements for first-time voters, the existence of various waivers for identification requirements at the polls, or even many other state-level regulations that should affect voter turnout.  I address these issues by conducting the first within-state analysis of the effects of voter identification laws on voter turnout using individual-level self-reports on voting from the Current Population Survey, the National Election Survey and the Cooperative Congressional Election Survey.  I also examine the differential effects of ID requirements on vulnerable groups, including racial and ethnic minorities, the elderly, and the poor.