Background: As a phased approach to initiating driving, graduated driver licensing restricts driving by young drivers with the aim of reducing crashes. It might increase riding with parents or on buses, which might be safer, or walking or biking, which might be more dangerous. We examined whether it increases nondriver injuries, and whether it reduces total injuries combining drivers and nondrivers. Methods: We conducted longitudinal analyses of 1995-2012 traffic injuries from 43 states. Using Poisson mixed regression, we estimated adjusted rate ratios for visible, incapacitating, and fatal injury. Results: Among 16 year olds, graduated driver licensing was associated with reduced passenger injuries (adjusted rate ratio 0.93, 95% confidence interval: 0.89, 0.97). It was not associated with increased injuries as bus riders, pedestrians, or bicyclists among 16- or 17-year olds. It was associated with a 10% reduction in total injuries among 16-year olds, but not 17-year olds. Conclusion: Graduated driver licensing was associated with reduced passenger injuries and total injuries.