Drivers with high crash risk such as teenagers have relatively low seat belt use, which contributes to their injuries. This article addresses whether belt use is lower in driving situations with higher crash risk (such as when teenagers transport passengers), further contributing to the problem. Belt use was investigated in a variety of situations among fatally injured drivers and their passengers, using 1995–99 data from the Fatality Analysis Reporting System. For drivers of all ages, belt use was lower at night and with alcohol present. For teenagers, but not adults, driver belt use decreased with increasing number of passengers. Driver belt use was lowest when teenage drivers were transporting passengers in their twenties, and many of these crashes involved driver alcohol use. Driver belt use was highest when teenage drivers were transporting passengers age 30 and older, many of whom could be parents. Passenger belt use also decreased among teenagers as the number of passengers in the vehicle increased. Transporting passengers up to age 29 should be restricted in graduated licensing systems, and restrictions should not be waived unless a licensed passenger older than that is present. Most existing passenger restrictions do not conform to this.