Most studies of adolescent work-related injury that have been published to date have been confined to administratively collected data available through state or national departments of labor and industry and/or occupational safety and health administrations. These data do not contain information on the severity or the sequelae of injuries. In addition, estimates of incidence may be biased due to inadequate data on the nature, amount, and seasonal variability of work done by adolescent workers. The present study is a cross-sectional survey of work and work injuries of 3,051 10th through 12th grade students from 39 high schools throughout Minnesota. Injury was defined as an event which caused any of the following: loss of consciousness, seeking medical care, and/or restricting normal activities for at least 1 day. The average hours of work per week during the summer and during the school year were 30 (95% Confidence Interval [CI] = 27, 33) and 16 (95% CI = 15, 17), respectively. There was no difference in hours worked between ethnic minorities and white students. The rate of reportable injuries was 12 per 100,000 hr worked (95% CI = 7, 18) for rural and 13 per 100,000 hr worked (95% CI = 7, 28) for urban females, 16 (95% CI = 9, 22) per 100,000 hr for urban males, and 20 (95% CI = 12, 28) per 100,000 hr for rural males. Ongoing medical problems were reported by 26% of the injured workers. Previous estimates of work-related injury to adolescents may have been low. More study is required to clearly define the incidence of injury and risk factors for these injuries.