Since nearly half of children fatally injured in automobile crashes were restrained, optimizing occupant protection systems for children is essential to reducing morbidity and mortality. Data from the Partners for Child Passenger Safety study were used to compare the differential injury risk between drivers and their child passengers in the same crash, with a focus on vehicle model year. A matched cohort design and conditional logistic regression model were used in the analyses. Overall, injury risk for drivers was higher than for children, but the risk difference was largest for the oldest model year vehicles, particularly for children aged 4-8 in seat belts. While drivers experienced significant benefits in safety with increasing model years, children restrained by safety belts alone derived less safety benefit from newer vehicles.