This paper examines the effects that downsizing has had on occupant injury. Statistical models are derived which demonstrate the relative risks associated with downsized cars and restraint use. Then actual occupant injuries are analysed to show how the total pattern of occupant injuries changes with downsizing.Each additional thousand pounds of vehicle mass decreases the odds of a driver injury in a crash by 34 percent when the driver is not restrained. For restrained drivers, this decrease is 25 percent per thousand. Restraint use further decreases the odds of a driver injury by two-thirds. To gain the same reduction in injury odds afforded the belted driver of a 2500 pound passenger car, the unbelted driver requires a 4325 pound car.For unrestrained occupants, the instrument panel, steering assembly and windshield (in frontal impacts) are the most frequent sources of injury. Restraint use reduces the chance of striking the windshield or instrument panel and increases the relative importance of the side interior surfaces as injury sources.