BACKGROUND: Physical abuse and assault are common problems in the Western hemisphere. The aims of this study were to investigate the injury incidence, distribution of injuries, the age and sex distribution, and the geographical differences in all patients admitted to Swedish hospitals between 1987 and 1994 because of injuries related to unarmed assault. METHODS: Patients admitted to hospitals in Sweden between 1987 and 1994 after physical abuse were included in the Swedish Hospital Discharge Register. A description of the types of injuries, surgical procedures, and lengths of hospital stay are presented. The change in the incidence of hospital admissions for unarmed violence-related injuries was evaluated. Linear regression analysis was used to correlate population density with incidence of hospital admission and to evaluate the change in age-standardized incidence of hospital admissions over time. RESULTS: Information was available on 17,453 persons, of whom 79% were males. The mean age was 30 years. Craniocerebral injury was the most common (72%) followed by injuries to the extremities (10%), thorax (5%), and abdomen (3%). The mean in-hospital stay was 3 days. Thirty-eight people (0.2%) died of their injuries. The age-standardized incidence of hospital admissions increased significantly over the years in males, but not in females. No correlation was detected between population density and incidence of injury. CONCLUSION: Young males are at the greatest risk of incurring physical injuries from assaults that warrant hospital admission, and the incidence in this group has increased significantly. Injuries to the head are the most common. Fatal injuries are rare. The in-hospital stay is usually brief. The frequencies of assaults are similar in urban and rural areas.