Purpose: Faced with an aging driving population, interest is increasing in the use of restricted licenses or "graduated delicensing" for older drivers to allow them to safely retain a driver's license. The primary purpose of this study was to determine whether restricted licenses are successful at mitigating number of crashes per year and whether they can extend the period of crash-free driving for aging adults. Design and Methods: Using a cohort study design, licensing and insurance claims crash records of all drivers aged 66 years and older in British Columbia were examined for the years 1999-2006. Nonparametric and Cox proportional hazards survival analyses were used to compare restricted vs. unrestricted drivers and to estimate crash risks. Results: The risk of causing a crash was 87% lower for restricted drivers compared with unrestricted drivers after controlling for age and gender. The most common restriction was a combination of daylight driving only plus a speed maximum of 80 km/hr. Restricted drivers retained a driver's license for a longer period of time than unrestricted drivers and continued to drive crash free longer than unrestricted drivers. There was no difference in severity of collisions, and results suggest a high level of compliance with daylight-only restrictions. Implications: These findings suggest that driving restrictions may be effective for prolonging the crash-free driving of some aging drivers, thus supporting their continued independence and delaying institutionalization. Further studies are needed to determine which drivers are most likely to benefit from restricted licenses.